Anda di halaman 1dari 242


This page intentionally left blank

W. D. Ross



Edited by
Philip Stratton-Lake


This book has been printed digitally and produced in a standard specification
in order to ensure its continuing availability

Great Clarendon Street, Oxford 0X2 6DP
Oxford University Press is a department of the University of Oxford.
It furthers the University's objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lutnpm Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan South Korea Poland Portugal
Singapore Switzerland Thailand Turkey Ukraine Vietnam

Oxford is a registered trade mark of Oxford University Press

in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York
© Oxford University Press 1930, 2002
The moral rights of the author have been asserted
Database right Oxford University Press (maker)
Reprinted 2007
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
And you must impose this same condition on any acquirer

ISBN 978-0-19-925264-0
W. D. Ross

I T is with much diffidence that I send into the world this essay on
two of the most important conceptions that occupy the attention
of moral philosophers, I am conscious of the great difficulty of most
of the main, problems of ethics, and of the fact that almost all the
assumptions that it seems most natural to make raise problems that
for their proper treatment, would demand long and careful discus-
sion, ! fancy that in some places, in ray anxiety to take account of the
complications and opposing considerations that demand to be taken
account of, I may have made the main outlines of ray view difficult
to follow; arid that in others, in ray wish to avoid undue complexity,
I may have made general statements without qualifications which
will suggest themselves as necessary, 1 have tried to strike a mean
between undue simplicity and undue complexity; but I cannot flat-
ter myself that 1 have always, or even usually, been successful. Some
of the conclusions I have reached seem to me almost certainly true,
and others seem to me very doubtful; and I have tried to indicate
which I think the more and which the less doubtful.
My main obligation is to Professor H. A, Prichard. I believe I
owe the main lines of the view expressed in my first two chapters to
his article 'Does Moral Philosophy rest on a Mistake?" (Mind, 1.912,
21—37), In addition to this, I have repeatedly discussed many of the
main ethical problems with him, and have learnt something from
every discussion; I have also had the advantage of reading a good
deal that he has written but not published. And finally, he has read in
manuscript most of what I have written, and has helped me greatly
by exhaustive comments and criticisms. These have been very prof-
itable to me both where (as in the treatment of lightness) he is (I
believe) in general agreement with my point of view, and where (as
in the treatment of the question what tilings are good) he to a large
extent disagrees.
I wish also to say how much I owe to Professor G. E. Moore's
writings. A glance at the index will show how much I have referred
to him; and I will add that where I venture to disagree, no less than
where I agree, I have always profited immensely from his discus-
sions of ethical problems,

October 1930,

Introduction fy Philip Siratton-Lake ix

A note on the text 1
Bibliography li



Editor's Notes 175

Index 179
This page intentionally left blank
Philip Stratton-Lake

Ross's The Right and the Good (1930) marked the pinnacle of ethical
inniiticmisra, a doctrine that had been the dominant moral theory in
Britain for much of the preceding two hundred years. But within
twenty years of its publication intuitionism had come to be rejected
out of hand by most moral philosophers, so that by the 1960$ It
became difficult to see what its appeal had been,' This negative
opinion prevailed for much of the second half of the twentieth cen-
tury, and meant that very few philosophers were inclined to read the
central, texts of intuitionism with any care or sympathy. This in turn
led to a caricature of intuitionism which encouraged the view that
intuitionism could easily and quickly be dismissed before moving
on to 'more serious* moral theories.
Ross's intuitionist moral theory suffered as much as others from
this neglect. Some critics maintained that his mtuitionism tells us
nothing that we did not already know, that he is content merely to
list our common-sense opinions,2 Others objected that his intu-
itionism is metaphysically and epistemologically extravagant in so
far as it assumes the existence of'otherworldly' non-natural prop-
erties that can be directly perceived by a mysterious special moral
faculty.* But all of this misrepresents Ross's views, and once the
caricature Is put aside, and we look at. Ross's 'actual views and argu-
ments, we discover a serious, honest, and persuasive moral philo-
sopher, whose views deserve far more serious consideration than
they often receive.

Thus, for example, in 1967 G, J, Warnock wrote that 'Inflationism seems, in retrospect, so
strange a phenomenon-...-a body of writing so acute and at the same time so totally unilluminat-
in.g~ that o.n.e may wonder how to explain it, what its genesis was' (Contemporary Mora!
Philosophy (MacMilian: London, 1967), i€).
See e.g. Mary Waraock, Etkics since ^oo (Oxford University Press: Oxford, 5968), 50, G. ].
Warnock, Contemporary .Moral Philosophy., 12—13, ant^ Alasdair Maclntyre, A Short History of
Ethics: A History of Moral Philosophy from the Homeric Age to the Twentieth Cwitury^ 2nd edn.,
(Routledge: London, 1998), 254.
William K. Frankena,/^^(Prentice-Hall: Etigiewood CHfis, NJ, 1963),86-^7.
The Central Theses
What, then are these views? In The Right and the Good Ross argues
for a distinctive view in the metaphysics of morals, in normative
ethics, and in moral episteraology. Ross is a moral realist. He
believed that tightness and goodness are objective features of the
world in just the way that shape, size, and mass are. *',.. there is', he
writes, *a system of moral truth, as objective as all truth must be*
(p. if), 4 and later on claims; 'The moral order,,, is just as much part
of the fundamental nature of the universe (and, we may add, of any
possible universe in which there are moral agents at all) as is the spa-
tial or numerical structure expressed in the axioms of geometry or
arithmetic' (pp. 29—•jo).5
We do not make actions right or good by our moral attitudes,
conventions, culture, or whatever, any more than we make some
thing square in these ways. Rather our moral attitudes are respons-
es to 'perceived '* goodness and tightness. And when our attitudes
are appropriate, it is because something is the way we think it is—
that is, it really is good, bad, right, or wrong. What makes our moral
belief that Xu good true (when it is true) is the property, or objec-
tive characteristic, X has of being good, according to Ross. In this
respect value judgements are like colour judgements.7 li my judge-
ment that the wall is red is correct, this will he because the wall lias a
certain colour property, die property of being red. Similarly, if my
judgement that something is good is correct, this will be because this
thing has a certain value property, the property of being good,
Ross's realism has the consequence that if there are no value
properties in the world-—no objective values-—then all of our value

Page numbers in parentheses in the text refer to The Riqkt and t/ie Good,
Here Ross commits himself not only to objective moral truth, but to necessary moral truths.
I put the word 'perceived' in inverted commas as Rons did not think f by? we Hterajly perceive
rightness and goodness. We judge that something is right or good not on the basis oi a perception
of these properties, but on the basis of a perception of other properties (The Foundations of Ethics
(Clarendon. Press: Oxford, ^939), 168).
Though it should be noted that there are many ways in which moral judgements differ from
colour judgements. For example, it is plausible to think that colour judgements are true if a cer-
tain subjunctive condition is truet namely, if we would have the relevant sensation under certain
specified circumstances, ft is not the case, liowever? that value judgements would be tme it it is
true that we would respond in a certain positive way under certain conditions. Ross notes certain
other differences between colour and value (pp. 88 and (21—2).
judgements will be mistaken. For ifit is the existence of moral prop-
erties that make our moral judgements true, then If there are no such
properties then none of our moral judgements would be true. To
think that all of our moral judgements are false is to embrace ari
error theoryJ about value.'
to the error theory,
J f
moral dis-
agreements are like disagreements about what colour something is
in an objectively colourless world. You might think that the ball is
red, whereas I might think it is orange. But we would both be mis-
taken. For we both think it has a colour which, in a colourless uni-
verse, it does not. The same would be true of disagreements about
the value of something in an objectively valueless world. You might
th ink that something is good, whereas I might think it is bad. But we
would both be mistaken. For we both think it has a value which, in a
valueless universe, it does not.
Although Ross's moral realism commits him to the view that in
an objectively valueless universe all of our evaluative judgements
would be mistaken (p. 82), he maintained that there is no reason to
think, that the universe is valueless (p. 82). His view is, then, that
some of our moral judgements are true, and what makes them true
is the presence of the relevant objective value.
Not. only was Ross a realist; he was a non-naturalist realist. This
means that he thought that moral properties cannot be understood
in wholly non-moral terms.8 Moral terms are evaluative or deontic
terms, such as "good", "valuable', 'ought*, 'duty', 'fitting', 'appro-
priate', or 'merited'. Non-moral terms are psychological, sociolo-
gical, evolutionary, or scientific terms, such as 'desire', 'approval*,
'society', 'survival', etc. For Ross, to understand moral properties
in wholly non-moral terms is to provide a reductive, naturalistic
definition of these properties.
If you define 'right* as meani ng what is approved by the community, you
are putting forward a naturalistic definition. If you define 'good* as
meaning "such that it ought to be desired", you are pulling forward a
non-naturalistic definition. Consequence theories also may be either
naturalistic or non-naturalistic. If you define "right' as "productive of the
greatest pleasure', you are putting forward a naturalistic definition. If

" The Foundations of Ethics, 6. We shall see later that tills understanding of the non-naturalness
of iiOR-naEural properties is inadequate.
you define it as 'productive of the greatest amount of good\ you are
putting forward a non-naturalistic definition.9
If moral properties can be defined in terras of natural properties,
then moral issues will be clecidable by means of empirical observa-
tion. If moral properties cannot be defined in this way, then moral
issues are not empirically clecidable. Ross does not deny that certain
empirical matters will be relevant to deciding whether we ought or
ought not to act in certain ways. His point is just that even when all
of the empirical facts are in, there is still a further, moral judgement
to be made—namely, that these empirical facts make a certain act
right, or good.
As well as insisting that tightness and goodness are non-natural
properties, lie held that they are simple properties. The view that
tightness and goodness are simple properties is the view that they
are not combinations of two or more properties or relations, such as
the relation of being appropriate (moral) and the property of being
approved (non-moral), or the causal relation (non-moral) and the
property of being best (moral). I0 Ross held that if a property is sim-
ple it cannot be defined. Consequently, he held that the simplicity of
tightness and goodness implies that these properties are indefinable.
In his normative theory, Ross defends a form of"methodological
intuitionism. Methodological intuitionists maintain that there is a
plurality of first principles that may conflict, and that no explicit pri-
ority rules for resolving such conflicts can be provided." This
means that principles of duty cannot ultimately be grounded in a
single foundational principle as consequentialists and Kantians
believe. This pluralism applies not only to the right, but to the good
also. There is, Ross maintains, no master value, such as pleasure,

Ibid., 6-7.
For Ross, a property is not complex simply in virtue of being a determinate tbrm of a deter-
minabte property. Vie did nor, for example, if sink that redness h'A complex property because h IKK
the (cJetenmnablc) property of being a colour and file (determinate) property of being red (p.
92 n.). In this he follows J, Cook-Wilson, Statement mdInference, ii(Clarendon. Press: Oxford,
1926), 502-4.
B. Williams, 'What does Intuitionism Imply?' in Williams, Making Sense ctf Humanity and
Othef Philosophical F\ipefs itySz—iym (Cambridge University Press; Cambridge, 599^}, ($2, Not
all pluralists are methodological tntuitionists, for a rationalized pluralism might have priority
rules. Methodological intnitionists iorro, therefore, a subclass of pluralist theories. In what
follows, however, I shall use the term 'methodological intuitionism* and 'pluralism* interchange-
well-being, or happiness, in terras of which the value of all other
goods can ultimately be understood. There are, rather, a number of
things that are good on their own account (Ross lists virtue, pleas-
ure, knowledge, and justice12) arid whose value does not, therefore,
need to be derived from the value of anything else,
Ross was an epistemological as well as a methodological intu-
itionist. Epistemological intuitionists maintain that we can know
directly (intuitively) that certain things are good or right on their
own account. We can know these things directly because the propo-
sitions which state that these things are good or right self-
evident. To say that these propositions are self-evident is to say that
they can be known, on the basis of an understanding of them, and so
can be known a priori. Ross did not maintain that all moral proposi-
tions are self-evident. Only the most fundamental moral proposi-
tions can be known in this way, according to Ross. Particular
judgements based on these foundations about what we should do in
specific situations are not self-evident, or even knowable. For Ross,
we can only have a probable opinion about such things.
We can, therefore, identify four key views which are at the centre
of Ross's intuitionist moral theory. The first is moral realism, the
second is the view that moral properties are simple, non-natural
properties, the third is his commitment to an irreducible pluralism
about the right and the good, and finally the view that fundamental
moral propositions are self-evident. The first two express his view
in the metaphysics of morals, the third expresses his normative
theory, and the fourth his epistemology. In what follows I consider
each of these aspects of Ross's view in more detail. In Section 11 dis-
cuss Ross's metaphysics of morals; in Section I I I focus on his nor-
mative theory—his pluralism about the right and the good; and in
Section HI I assess his moral episteraology.

By justice Ross means the allocation of happiness, or pleasure, in proportion to merit
(virtue). He does not always mention this value when he lists the bask goods, and so sometimes
says that there are three things that are intrinsically good. IE is clear, however, that he thought that
justice is a distinct intrinsic good which differs in structure from the other three in that it is com-
plex, while they are simple.
Moral Truth and Moral Realism
Why should we think that there are moral as weil as non-moral
properties in the world? Why think that values are sewn into the
fabric of the world, as it were? Ross claims that moral realism is built
into our everyday ethical discourse and thought. For example,
although we can only discover the value of certain things by attend-
ing to them, we think of them as having this value before we attend-
ed to them, and as possessing this value evert if we did not attend to
them (p. 81), We think that the positive feelings we have towards
those things we judge good are responses to value rather than as
conferring value on their objects (p. 89), and we think of moral
judgements as expressing moral beliefs which aim to track objective
evaluative facts (p. 85),
More generally, Ross thinks that simply in virtue of being an
adjective there is an assumption that 'good' refers to a property,
"When some entity is commonly referred to by an, adjective, there is
a certain presumption that it is a quality, just as when it is common-
ly referred to by a prepositional phrase, there is a presumption that
it is a relation or a relational property,' 1 * When we think that some
adjective applies to something, we naturally think, of that thing as
having a certain feature, of being a certain way; and to think of it as
being a certain, way is to think of it as having a certain property. This
presumption may, on reflection, be overturned. 14 It may turn out
that the adjective 'good' refers to a relation, or relational property,
rather than an intrinsic property of the things of which it is predicat-
ed, It may even turn out that in judging that something is good we
are saying nothing at all about that thing, but are either simply
expressing our positive attitudes towards it, or are saying that such
attitudes would be appropriate. But Ross rejects these alternatives
and insists that there is no good reason to think that the presumption
that 'good* and 'right* do not refer to properties of the things of
which they are predicated (p. 114).'' The realist construal of moral-
* The Foundations of Ethics, 278. See also The Right and tke Goad, 81 -
Ross was an irrealist about colour even though Veef is an adjective (p. 8<S).
See The F&umia.tiQn^ of Ethics,^ f£, for Ross's rejection of Carnap's theory, pp. 34 il. for his
critique of Ayer's emotivism, and pp. 279 ft", for his response to Brentano's claim tiiat "good' is a
Sch&insquah tiii.
if.y is, therefore, the best way of capturing the moral phenomenolo-
gy and what might be called the platitudes oi ordinary moral
Recently, however, philosophers .have denied this. They accept
that the platitudes of ordinary moral thought give us good reason to
think of moral judgements as truth apt—that is, as being either true
or false-—but deny that truth in ethics commits us to any sort of real-
ism. They do this by claiming that moral judgements are true (or
false) only in a minimal sense.16 According to this minimal sense
one says no more in saying * "Rape is wrong" Is true' than if one had
said " Rape is wrong*. To add 'is true' to a quoted sentence is just to
note the appropriateness of disquoting. It does not imply that reali-
ty corresponds to the way the sentence represents it as being. Minimal
truth in ethics does not, therefore, imply that the world is a certain
way, as realists like Ross think, but implies simply that one has moral
commitments.* 7
But as Jackson, Oppy, and Smith have pointed out,18 the mini-
malist conception at truth is inadequate because it fails to pick out
what it is about a certain discourse that makes it truth apt. Mot
everyone agrees that moral judgements have a truth value.
Emotivists think that moral judgements express non-cognitive atti-
tudes rather than beliefs, and these attitudes cannot be true or false.
So 'why should we think that moral discourse is truth apt, and that
emotivism is mistaken? Moral discourse is truth apt, minimalists
maintain, if certain syntactical and normative conditions are met.
The syntactical conditions are that sentences be significantly
erabeddable in suitable constructions, e.g. negation, conditional,
prepositional attitude, and truth ascriptions. The normative
conditions are that there are certain well-established standards in a
community governing the use of such sentences. Since these syntac-
tical and normative conditions are easily met, minimalists have no
trouble allowing that moral sentences (as well as sentences within
See e.g. S. Blanckburn, Ruling Passions: A Therpry oof Practical Readson ( Clarendon Press:L
Oxfor, 1998).
NM.Smith, "Moral Realism', in H.Lafoollette(ed.) Ethical Therory (Blackwell:Oxford,200),
F. Jackson. G. Oppy, and M. Smith, 'Minimalism and Truth /Vptness', Mind, 103 (^994),
287-502. See also Smith, 'Moral Realism", 15—37, for an excellent account of minimalism about
trutli and the problem with it.
other problematic discourses) are truth apt. But since these condi-
tions only quality a discourse for truth, in the minimalist sense, truth
aptness does not imply any problematic realism,
The problem with the minimalist criterion of truth aptness, how-
ever, is that it is arbitrarily selective about which platitudes it allows
to determine how we are to understand truth aptness. One of the
platitudes of a truth apt. sentence is that it be used to give the content
of a belief of someone who asserts the sentence.19 This condition is
neither syntactical nor normative, but is, nonetheless, central. If,
however, it is included amongst the platitudes determining truth
aptness, then truth aptness will no longer be a minimal matter. For
beliefs are states that are designed to fit the facts, are sometimes
action-guiding, and tend to evolve in a rational manner.20 Whether
the utterance of certain sentences expresses such a state is a substan-
tial rather than a minimal matter. It is not, therefore, clear that
we can admit that moral discourse is truth apt without realist
assumptions, as minimalists maintain,21
Simplicity and Non-Naturalism
But even if moral discourse is truth apt in a fairly robust sense—that
is, in the sense that moral sentences are true if and only if the world
is tile way they represent it as being—Ross's moral realism might he
mistaken. We might have good reason to think that there are no such
things as moral properties. Moral properties might be metaphysi-
cally weird, or have no place in a modern scientific view of the
world. If this were the case, then the conditions for the truth of our
moral judgements would never obtain, and an error theory would
be correct The question we have to address now, therefore, is
whether we have good reason to think that there are no such things
as moral properties.
We would have no problems accepting that moral properties
exist if these properties are natural properties, for no one thinks that,
natural properties are peculiar or incompatible with the scientific
Weltanschauung, It is for this reason that most contemporary moral
n 20
' Jackson etaL, "Minimalism and Troth Aptness', 294. Ibid. 296.
' Minimalists could insist that such utterances express beliefs minimally construed. But if this
is to be pursued, then minimalists owe us an account of what makes a state one of belief rather than
something else (ibid. 296-7).
realists are naturalists. Ross, however, like a!J inflationists,, Thought
that moral properties must be non-natural properties, and (like
many other intuitionists) that rightness and goodness are simple
non-natural properties,
Two questions arise at this point, (i.) Why did Ross think that
rightness and goodness are simple non-natural properties? (2) Do
such properties exist? If Ross can provide us with no good reason
for thinking that rightness and goodness are simple, non-natural
properties, then the view that such properties exist 'would be urtcon-
troversial, for no one denies the existence of natural properties. If,
however, Ross can provide us with good reason to think that truth
in ethics depends on the existence oi: simple, non-natural properties,
then the view that there are such properties will be controversial, as
many philosophers deny the existence of non-natural properties. If
they are right, and if moral truth depends on the existence of non-
natural properties, then an error theory will be true.
Most discussions of die intuitionist view of moral properties do
not clearly distinguish between the claim that goodness and right-
ness are simple, and the claim that they are non-natural properties.
These are, however, very different and independent claims. A nat-
uralist might agree with Ross that rightness and goodness are sim-
ple properties, but insist that they are simple natural properties, or a
non-naturalist might insist that rightness and goodness are complex
properties, but agree with Ross that they are non-natural,22 It is, I
believe, important to note this distinction, for if we do not, then it
will not be clear whether a particular problem is generated by the
simplicity of rightness and goodness, or from their non-naturalness.
! shall, therefore, deal with these separately.
Why does Ross think that goodness and rightness are simple
properties? He rests his view on two closely interconnected
arguments. The first may be called the 'transparency argument*.
According to the transparency argument, if 'good* could be
defined, the term would have to stand for a certain complex proper-
ty, for the definition would consist in specifying what this complex
A. C. Ewing argued that goodness is a complex non-natural property (The Definition of
Good (Rotitledge & Kegan Paul: London, 1947), ch. 5), and in Pmuipia Etkica (rev. edn.
(Cambridge University Press: Cambridge, 1995), 196) Moore argued that tightness is complex,
but non-natural.
Is. If'good* stood for a complex and could thus be defined, then we
could only use the term "good' Intelligently and intelligibly to die
extent that we had this definition in mind. But we use this terra intel-
ligently and intelligibly without having any definition in mind.
Therefore, 'good* and the property this term, stands tor must he
simple (p. 92),
The second argument Ross proposes is Moore's open question
argument. According to this argument every attempt to define
"good" must fail, for it is always possible to enquire whether some-
thing that answers to the supposed definition is good. We can show,
for example, that being good is not the same as being desired by
somebody, for even if we know that war is desired by someone, we
may still doubt whether it is good (p. 92), If the definition were cor-
rect, however, we could have no such doubt. One could not, for
example, intelligibly doubt that John is a bachelor if one knew that
John is a male and is unmarried. Since such doubts are always pos-
sible in relation to any definition of'good', all such definitions must
be false. Consequently, 'good * must stand for a simple property.
Although these are different arguments, Ross regards the second
as adding nothing to the first.25 He recognizes, however, that the
transparency argument cannot be accepted as it stands, for, he
notes, we do not always have a certain complex property in mind
when we use a terra that denotes a certain complex,
... there seem to be cases in which we seek for the definition of a terra and
finally accept one as correct. The fact that we accept some definition as
correct shows that the term did somehow stand for a complex of ele-
ments; yet the fact that we are for some time in doubt whether the term is
analysable, and if so, what the correct analysis is, shows that this complex
of elements was not distinctly present to our mind before, or during, the
search lor a definition. It appears as if we cannot avoid recognizing that
there is such a thing as using a term which implicitly refers to a certain
complex, while yet the complex is not explicitly present to our minds . , ,
(pp. 92-3)
This acknowledgement seems to be fatal to Ross's argument for the
simplicity of moral properties. If a terra may be definable even
He says they both express the same thought, namely, that if good stood for a complex we
would have this complex in mmd when we use the term. Yet we have no such complex in mind
when we use the term 'good1.
though we do not have this definition in mind when we use the term,
then the term 'good' could be definable (complex) despite the fact
that we do not have this definition in mind when we use die term,
and so may intelligibly question whether something that satisfies
the definition is good. But Ross insists that this opacity only shows
that we cannot know straight off, as it were, that all definitions of
"good* must fail. 'What we roust do is attend to any definition that
seems at all plausible, and ask ourselves whether this is 'what we
meant by 'pood' all along (p. 93). If the answer to this question is
'no', then we should reject the proposed definition. And if we reject
all of the most plausible definitions of'good* in this way "we may
feel fairly confident that "good" is indefinable * and stands for a sim-
ple property (p. 93), Ross's point is that even if good is an opaque
complex notion,, once the definition is presented to us we should be
able to respond by saying, 'that is what 1 meant all along, though I
was unable to formulate it'. Applying this method to both tightness
and goodness, Ross concludes that they are both simple properties.
Ross thought, that goodness and Tightness are not only simple,
but non-natural properties. If'good' and "right* cannot be defined,
then they cannot be defined in wholly natural (non-moral) terras.
But even if goodness and tightness cannot be defined (are simple
properties), they may still be natural properties. 24 Ross did not
clearly separate the question of whether moral properties are simple
from the question of whether they are natural. As a result, of this he
offers no argument against the view that moral properties might be
simple natural properties, In one respect his failure to distinguish
these issues is unproblematic, for it is difficult to imagine any simple
natural property that might plausibly be identified with goodness
(or lightness). And even if some simple natural property could be
found that could plausibly be identified with goodness (or right-
ness), Ross could use the same argument against this identification
that he did to establish the simplicity of goodness. He could claim
we should consider whether we have always had. this simple natural
property in mind when, we think of something as good (or right).
Although we cannot know for certain beforehand that all such
"' Ho&s held that complex properties presuppose simple ones (p. 93). Since he does not seem
to think that the natural realm is reducible to some other realm, he is committed to there being
simple natural properties.
Identifications must fail, the absence of any plausible candidate
property means that it is highly unlikely that goodness is identical
with some simple natural property. If, therefore, naturalists are
right, then goodness and Tightness are identical with some complex
natural property- Consequently, if Ross is right to reject all defini-
tions of tightness and goodness, then he is right that the moral can-
not be reduced to the non-moral (natural),
I do not, however, think that Ross is right that all definitions of
tightness and goodness are inadequate. Ross thinks that the open
question argument illustrates the fact that we do not have some def-
inition in mind when we think of something as good. But I think this
argument shows something different. I think it shows the essential
normativity of evaluative judgements.2* To say that evaluative
judgements are normative is to say that they express conclusions
about what would be reasons tor acting or responding in certain
(positive) ways. Natural facts may provide the ground for such
practical conclusions, but judging that these natural facts obtain
does not by itself involve explicitly drawing these conclusions. It is
for this reason, therefore, that questions such as 'This is N, but is it
good?' (where W stands for some natural property) have an open
feel. They have an open feel because when we decide that some-
thing is good we conclude that we have reason to adopt some pro-
attitude towards it. But in judging simply that it has some natural
property, we have not yet concluded that we have reason to adopt
such an attitude towards it.26
This is not to say that natural properties cannot give us reason to
respond in certain positive ways. On the contrary, the fact that
something will give another person, or me, pleasure gives me a rea-
son to get it, either for the other person or for myself. The point is
that the judgement that I have reason to respond in a certain way is
always a further judgement over and above the judgement that the
natural property obtains, It is one thing to judge that something will
be pleasant, and quite another to judge that its pleasantness gives me
a reason to pursue it.
! here follow T. M, Seankm's account of the open question argument and the eotichision he
draws from this in What We Owe to Each Qtli&r (Harvard University Press: Cambridge, Mass.,
1998), 9« ff.
" Ibid. 96.
But the judgement that we have reason, to respond in a certain
positive way to something does not seem to be a judgement over and
above the judgement that this thing is good. On the contrary, it
seems that in judging that something is good we are judging that we
have reason to respond in certain positive ways.27 If this is right,
then it strongly suggests that thinking of something as good is
thinking of it as having features that give us reason to respond in
certain ways (to adopt some pro-attitude or to do some action). The
open question argument should not, therefore, lead us to conclude
that goodness is indefinable, but to the conclusion that a certain def-
inition is correct- namely one in terras of reasons, pro-attitudes,
and actions, But although the open question argument does not sup-
port the view that goodness is a simple property, it does, 1 believe,
support the view that it is a non-natural property. For what this
argument makes clear is that goodness is a normative property
whereas natural properties are not. If they were, then we would be
able to judge that we have reason to adopt some pro-attitude
towards something simply in judging that it has some natural prop-
erty. Since the judgement that we have such reason is always addi-
tional to the judgement that something has a certain natural
property, natural properties are not normative.
Someone might insist that the property of being a reason is a nat-
ural property. But I think no naturalistic account of reasons can be
accepted. The most plausible ones try to define reasons in terms of
what we would desire under certain conditions. But either some
explicitly normative term will appear in the list of conditions-—in
which case the analysis will not be reductive—or the analysis is
reductive, in which case it will be vulnerable to a version of the open
question argument. For we can always doubt that something gives
us a reason even though we know that the reductive definition is
What we get from the open question argument, then, is not that
goodness is indefinable, but that it is indefinable in naturalistic, or
non-normative, terms,28 The open question argument actually

Ibid, 97 (if.
This was in fact what Moore was most concerned to show whli the open question argument
[Pnncipig Etkicn^ 19).
favours a certain non-naturalistic definition of goodness-—it makes
clear that for X to be good is for X to have features that give us rea-
son to adopt some pro-attitude towards it,
Since the open question argument can be applied to tightness as
well as to goodness, it would seem that we can draw similar conclu-
sions about lightness-—that is, we can conclude that tightness can-
not be understood in wholly natural terms, and that it should be
understood in terms of reasons. But since, as Ross rightly notes,
tightness (where tightness is understood as equivalent to ought-to-
be-doneness) is distinct from goodness, the property an act has of
being right cannot be the property it has of having features that give
us reason to adopt some positive attitude to it. It is, rather, the prop-
erty an act may have of having features that gi\re us overall reason to
do it. It is for this reason that questions of the form 'Action A Is A-7,
but is it right?' ('where '/V* stands for some natural property) have an
open feel. They have an open feel because in asking whether an
action is right (whether we should do it) we are asking whether we
have overall reason to do it, and although ,/Vmay give us overall rea-
son to do A, we do not conclude that we have such reason simply in
judging that the action is N. There is, however, no question of
whether we have overall reason to do A once we have concluded
that A is the right act. This is, i believe, best explained by identify-
ing Tightness with the property A\a& of having features that give us
overall reason to do it.
So I think that Ross was wrong to conclude that Tightness and
goodness are simple properties, and that 'right' and 'good* are inde-
finable. But Ross does not need to insist on the indefinabili ty of these
notions to maintain that tightness and goodness are non-natural
properties. For although pro-attitudes and actions are natural prop-
erties, the property of being a reason is, I believe, a non-natural
property. The property of being a reason, is a normative property,
and one that cannot be understood in 'wholly non-normative (natu-
ralistic) terms. This point is obscured by Ross's characterization of
non-natural properties as those which cannot be understood in
wholly non-moral terms. For the term 'reasons* is neither an explic-
itly moral nor a natural term. So once we note that reasons are non-
natural properties, we have to characterize the natural—non-natural
distinction in some other way, I do not know of a wholly adequate
INTRO D U CT1 ON xxiii
characterization of this distinction,29 but Moore captures the flavour
of this distinction when he says that natural properties are those that
are the business of the empirical sciences (including psychology) to
study, and non-natural properties are those that cannot be studied
empirically. This characterization of the distinction allows us to
capture the view that die property of being a reason, is a non-
natural property. For although science can tell us, for example, that
certain, animals feel, pain, it cannot tell us that this empirical fact
gives us reason not to do certain things, say, boil them alive. In point-
ing out that certain animals can feel pain, most scientists will have
little doubt that this empirical fact gives us reason not to boil them
alive. But their belief that we have such reason will not be the result
of scientific investigation. All such investigation, reveals is the non-
normative (empirical) fact that a certain animal can. feel pain. It does
not reveal the normative fact that this empirical fact gives us reason
to act in certain ways.
So far, I have argued, Ross has provided us with reason to think
that moral properties are non-natural properties, but no good rea-
son, to suppose that they are simple. But as we noted earlier, many
philosophers fee! that if the truth of moral judgements commits us
to the existence of non-moral properties, then we must abandon our
belief that some moral judgements are true and accept an. error
theory. For non-natural properties are unacceptably weird and
mysterious. Now the question of whether such properties are as
metaphysically suspect as naturalists insist raises deep and complex
issues that cannot be gone into here. "We can, however, make a
number of brief points on this topic.
First, we may put aside two objections that are based on contu-
sion. Sometimes the weirdness of non-natural moral properties is
expressed by saying that they are otherworldly properties. But Ross
did not think that moral properties exist in some other world. He
doesn't think that: certain things in this world are good because in
Thomas Baldwin suggests that we understand non-natural properties as properties that lack
causal powers (introduction to Principia Ethica, p, xxii). Although I think that this, tike Moore's
definition, is helpfel in characterizing tbe natural—non-natural distinction, I do not think it can be
accepted as a definition.. For some philosophers think that dispositionai properties, such as fragili-
ty or solubility lack causal efficacy. If they are right, then all dispositional properties woold be
non-natural properties, according to Baldwin's definition. But fragility and solubility are clearly
natural properties, whether or not they have causal powers.
some other world goodness exists. Moral properties are ways in
which natural things in this world are. Since the way in which some-
thing Is cannot be separated from the thing that is in that way, if
moral properties are non-natural properties of'naturai (thisworid-
ly) things, then moral properties must be thisworldly also,
A second common objection to Ross's view about moral proper-
ties is that they are mysterious because nothing can be said about
them,10 This objection is confused in a number of ways. First, it has
nothing to do with the supposed non-naturalness of moral proper-
ties. If it is true that moral properties are mysterious because
nothing can be said about them, then it is true not because these
properties are non-natural but because they are indefinable.
Second, it is not clear that: a property is mysterious simply in virtue
of being indefinable. For Ross, a property is indefinable if and only
if it is simple, but there is nothing mysterious about simple proper-
ties as such, I take it that no naturalist will claim that the basic ele-
ments of die material world are mysterious just because they are
simple. Finally, Ross did not think that nothing could be said about
simple, indefinable properties. All their simplicity roles out is a
definition. It does not prevent there being many informative, non-
definitional things that can be said about such properties,31 So in so
far as this criticism has any force at all, it does not count against the
view that there are non-natural properties, but against the view that
there are .simple moral properties, and it does not have much force
against this view. In any case, I have claimed that Ross's arguments
only support die Yiew that moral properties are non-natural prop-
erties, and actually support the view that they are certain complex
properties. If I am right, then the claim that such properties are mys-
terious because they are simple is irrelevant (as well as confused),
A third, common objection is epistemological. It is that, non-
natural moral properties could only be known by means of a special
moral faculty, But such a moral faculty is as mysterious as the
properties it is supposedly able to detect. There is a real problem
here for non-naturalists like Ross, but it is not at all clear that he can-

™ Warnockj Conwnporary Moral Philosophy,, n.

In Tke Foundati&m of Ethics Ross claims tliat moral fittmgncss is indefinable, but points out
what it has in common with utilitarian and aesthetic ftttingness, and how it differs from these (pp.
5' ff.).
INTRO D U CT1 ON xx. v
not provide a decent response. First, once we recognize that non-
natural properties are, for Ross, thisworldly, our knowledge of
them will seera less mysterious, for it will be knowledge of this,
rather than of some other, world. Secondly, as we shall see later,
Ross did not think that we know that certain things are intrinsically
good or prima facie right literally by perceiving the property of
goodness or prima facie tightness certain things have, let alone by
perceiving them by means of a special faculty. We know these things
by reflection on certain moral propositions, and such, reflection
requires nothing more than the ability to know anything by a priori
reflection—that is, the ability to understand and think. l2
Other objections to non-natural properties are, however, harder
to deal with. But if we abandon die view that tightness and goodness
are simple non-natural properties and understand them in terms of
reasons, it: is much harder for worries about non-natural properties
to lead one to adopt an error theory about value. If moral properties
are simple, then pervasive error in our evaluative judgements will be
relatively local. For we can doubt the existence of simple moral
properties without doubting anything other than, those properties
and the complex, moral properties that are formed out of them. If,
however, value is to be understood in terms of reasons, then (assum-
ing one lias no problem with actions and psychological attitudes)
one's scepticism about value would have to stem from scepticism
about reasons. But if an error theory about value rests on an error
theory about reasons, then the error theorist would be committed
both to the view that we have good reason to deny the existence of
reasons, and that: this implies that all of our reasons judgements are
false. But such a view seems to undermine itself, for it is to say that
we have reason to be sceptical about reasons, and implies that it is
false that we have reason to be sceptical about reasons.
One might try to avoid this conclusion by limiting one's scepti-
cism to practical reasons—that is, reasons to do some act or adopt
some pro-altitude, This form of localized scepticism would leave
epistemic reasons (reasons to believe) untouched, and so would not
land the error theorist with a self-undermining view. But I do not
see how scepticism about reasons could, be limited in this way. The
See L. BonjourT In Defence &f Pure Reason: A Rationalist Account of A Privri Justification
(Cambridge University Press: Cambridge, 5998), IQJ-~<).
difference between practical, and epistemic reasons is not. that they
stand in different warranting relations to certain things (one rela-
tion queer and the other Innocuous), but that they warrant different
things. Practical reasons warrant pro-attitudes and actions, where-
as epistemic reasons warrant beliefs, li, therefore, one lias doubts
about: the normative (warranting) relation itself, these doubts could
not he localized in such a way as to avoid scepticism about epistemic
as well as practical reasons,
If the above argument is along the right lines, then the naturalist
must abandon either the view that moral discourse is truth apt-—the
option taken by emotivists- or deny that its truth aptness implies
the existence of non-natural properties—the option taken by natu-
ralist realists. But I believe that we have good reason to think of
moral discourse as truth apt, and Ross offers a persuasive argument
that moral properties (the truth-makers of moral judgements) can-
not be identified with, any natural property. My view is thus that we
should abandon naturalism.

Critique of Consequentialism
Intuitionists are often accused of being obsessed with defining
'right' and 'good*. But Ross clearly thought (wrongly, I have
argued) that nothing interesting could be said on this subject, for he
thought that goodness and tightness are simple and thus indefin-
able, Ross's concern with the definition of tightness and goodness
was purely negative, He clearly thought that all definitions of good-
ness and tightness are mistaken, and that this mistake had implica-
tions for our thinking about normative ethics. For if one thought, as
Moore did in Principia Ethica* that to think of an act as right is to
think of it as maximizing good outcomes, then consequentialism
must be correct, despite the fact that we often do not seem to look to
future consequences when considering what we should do. But as
Ross points out, and, Moore eventually came to see,31 if consequen-
There is evidence that Moore recognised this point in his Ethics (Oxford University Press;
London, 1966). For a discussion, of this, see Ross, pp. 8—u. This recognition is explicit, however,
in his 'A Reply to my Critics', in P. A. Scliilpp (eel.). The Philosophy ojG.E, Moan (Northwestern
University Press: EvaiisEOK, III., 1942), 5^8.
INTRO D U CT1 ON xxvii
tiatism is true, then it would not be analytically true, but would be a
substantive (synthetic) truth. In so far as consequentialisra is at all
plausible, it does not tell us what it is for an act to be right, but offers
an account of the ultimate reason why all right acts are right—that
is, the ultimate ground ot tightness. If consequentiatists cannot force
their doctrine through by insisting that it is analytically true, then
the case for consequential ism must be made by considering differ-
ent types of right action and considering what makes these actions
right (p. '56). Ross argues that once consequentialists are forced to
defend their view in this piecemeal fashion, they become vulnerable
to a series of objections. These objections may not be thought to be
individually decisive, but together they do form a compelling case
against consequentialism.
First, Ross thinks consequentiaiism oversimplifies the morally
relevant relations in which we stand to others, Consequentiaiism,
Ross maintains, implies that 'the only morally significant relation in
which my neighbours stand to me is that of being possible beneficia-
ries by my action' (p. 19), But, he objects, although the benefac-
tor—beneficiary relation is morally important, it is not the only
morally relevant, relation. The relations of promiser to promisee, of
creditor to debtor, of wife to husband, of child to parent, of friend to
friend, of fellow countryman to fellow countryman are also moral-
ly relevant to determining what we should do (p, 19).
Ross, however, overstates his case here. Consequentiaiism does
not imply that the only morally significant relation in which we stand
to others is that of beneficiary to benefactor. Consequentialists can
allow that all of the relations Ross mentions are morally significant
also, Consequentialists do not claim that these relations are morally
insignificant, but that their significance is derivative, and contingent
on the fact that keeping one's promises, paying one's debts, etc.
happens to produce the best outcome. But a consideration that is
morally relevant under certain conditions is morally relevant; when
those conditions obtain. Ross's objection must, therefore, be that
these other relations are morally significant In their own rig/it^, and
thus do not need to acquire this significance because luck would have
it that respecting these relations maximizes good.
But why should we agree with Ross that these considerations are
morally relevant in their own right and reject: the consequentialist
account of their relevance? Ross offers a number of arguments to
suggest that being productive of the best outcome is not the only
basic moral consideration. The first is that the fact that we have
made a promise can decide what we should do where breaking our
promise will produce as much good as keeping it.
If, so far as I can see, I could bring equal amounts of good into being by
fulfilling my promise and by helping some one to whom I had made no
promise, I should not hesitate to regard the former as my duty. Yet on the
view that what is right, is right because it is productive of the most good I
should not so regard, it. (p. 18)
If all moral considerations acquire their moral significance from.
consequence-based considerations, then if the value of the conse-
quences of keeping a promise is the same as the value of the conse-
quences of breaking it, we should have no moral reason to keep our
promise rather than break it. But we have no doubt that in such a sit-
uation we should keep our promise rather than break it. This can
only be explained by the fact thai promise-keeping has a moral
importance that does not stem from the value of outcomes.14
Ross's second argument is that consequentialism ignores what he
calls the 'highly personal character' of duty (p. 22),
If the only duty is to produce the maximum of good, the question of who
is to have the good whether it is myself, or my benefactor, or a person
to whom I have made a promise to confer that good on him, or a mere fel-
low man to whom I stand in no such special relation—should, make no
difference to my having a duty to produce that good. But we are all in fact
sure that it makes a vast difference, (p. 22)
The thought here seems to be that we are more certain that these
relations make a difference to whom we should confer the good than
we are of the truth of consequentialism. If this is correct, then if con-
sequentialism implies that these relations make no such difference,
then we should reject consequentialism rather than our conviction,
that these relations make a difference.35

Brad Hooker addresses this son of difficulty as h applies to rule consec]uenlialism in his Ideal
Cade* Real World (Oxford University Press: Oxford, 2000), H4™ 16.
This argument from degrees of confidence is pervasive in Ross, and bears a dose affinity to
Moore's argument against scepticism about the external world in his 'Some Judgements of
Perception', in Philosophical Studies (Kouriedge & Kegari Paul: London, ^922).
Ross also notes that we do not think that, the tiniest increase in
overall 'value would necessarily justify us in breaking a promise, if,
for example, ray keeping my promise to A would produce 1,000
units of good for him, but by doing some other act I could produce
i. ,001 units of good lor 2?, I would not think that I should break my
promise to A (p, 34).3<s "To produce die 1,001 units of good for B
rather than fulfil on r p remise to A would be to take, not perhaps our
duty as philanthropists too seriously, but certainly our duty as mak-
ers of promises too lightly* (p, 35).
Ross goes on to offer two variants of the same thought experi-
ment. The first involves producing 1,001 units of good for A by
doing something other than what I promised, which would produce
i,000 units of good for A. The other is where I have made no
p romise, but have the option of producing i ,000 units of good for ^,
who is a very good man, or 1,001 units of good tor B., a very bad
man. In the first case we should, Ross claims, do what we promised
A^ while in the second we should produce the benefits for A. But in
both cases what we should do is produce the lesser good.
Ross further that, 'if it were self-evident that the right
coincides with the optimific, it should be self-evident that what is
prima fade right is prima. facie optimific. But whereas we are certain
that keeping a promise is prima facie right, we are not certain that it
is prima facie optimific* (p, 36). What each of these arguments
aims to do is to show that on reflection we do not think that consid-
erations of fidelity are morally relevant just because they are
instrumental to bringing about the best state of affairs. These con-
siderations have intrinsic deontic relevance which, as such, does not
need to be inherited from anything else, The duty to keep our
promise may be overridden if breaking it would bring about a con-
siderably better state of affairs. But the fact that considerations of
fidelity can be defeated by consequence-based considerations does
not imply that when we should keep our promises this is also
explained by consequence-based considerations.

Ko&s later considers a response to tltis objection, that failing to keep one's promise will liave
effects that will always tip the evaluative balance in such cases in favour of keeping one's promise
to A (p. 38), but riglltly refects tills.
Other Attempts to Derive the Right from the Good
Consequentialists maintain that tightness Is grounded in good-
ness- -that the ultimate reason why some act is right Is because it is,
In a certain sense, good. Consequentialisra is not, however, the only
theory that attempts to ground the right in the good. Certain theo-
ries attempt to ground the tightness of right acts, not in the value of
their outcomes, their instrumental goodness, but in their intrinsic
moral value. On this view, we ought to do certain acts because those
acts are morally good on their own account. Egoists also attempt to
ground the right in the good. For them we ought to do certain
actions because doing these acts will, in the long run, promote our
own good. Ross quickly disposes of both of these attempts to derive
the right from the good. The egoistic theory is rejected on the
ground that 'a great part of duty consists in an observance of the
rights and a furtherance of the interests of others, whatever the cost
to ourselves may be' (p. 16), Furthermore, even if it were true that
doing morally right acts always benefits the agent in the long run,
this would be irrelevant to the moral Tightness of those actions, 'As
soon as a man does an action because he thinks he will promote his
own interests thereby, he is acting not from a sense of its tightness
but from, self-interest* (p, rf>).
Given that Ross thinks that the tightness of an action is com-
pletely independent of the motive from, which it is done (pp. 4 ft".), it
is not cl ear how this point about acting from different motives is rel-
evant. But I take Ross's point to be this. When we think of ourselves
as doing the morally right act, we regard ourselves as responsive to
considerations such as fidelity, intrinsic value, gratitude, etc. 'We do
not, however, think of self-interest as a moral consideration—that
is, as one that could generate a moral obligation. So when we think
of ourselves as responding solely to considerations of self-interest,
we do not think that we are doing'what we morally ought to do. The
'ought* that considerations of self-interest generate is not a moral
ought. It is for this reason that considerations of self-interest are
irrelevant to the moral Tightness of an act. What is relevant are the
considerations that can make an act morally right, but considera-
tions of self-interest will not appear on a list, of such considerations.
Consequently, considerations of self-interest are irrelevant to the
rightness of morally right acts even if all morally right acts tend to
promote the agent *s Interest.
Ross spends a little more time on the attempt to derive the right
from the morally good. He does this in the chapter on the meaning;
of 'right', and so does not explicitly consider this view as a theory ot
fae ground of rightness. But by arguing that what is morally good is
never morally required he tries to show that 'morally good* need
not be coextensive with 'morally right'. If this attempt is successful,
it would show not only that 'right* cannot mean 'morally good', but
also that .moral goodness cannot be the sole ground of morally right
Ross's arguments against the view that we are never required to
do morally good actions rest on the assumption that the only moral-
ly good, actions are those that are done from morally good motives
(p. 4), Granted this assumption, ii he can show that we are never
required, to act from a morally good motive, then he will have
shown that we are never morally required to do a morally good act,
In The Right and the Good Ross proposes two arguments for the
view that we are never morally required to act from morally good
motives,37 His first argument is that such a requirement would
transgress the 'ought implies can' principle, His second argument is
that the requirement to act from the morally best motive (the motive
of duty) leads to an infinite regress.
We cannot be required to act from a good motive, because we can
only be required to do what we can, at a moment's notice, choose to
do. But we cannot choose to have a motive if we do not have it, and
even if we possess the relevant motive, we cannot choose to be moti-
vated, by it rather than by some other motive. The object of choice,
what we choose, is some action, and although we choose to do this
action from some motive, the motive from which we choose it is
never part of what is chosen,*8 Since we cannot choose to be moti-
vated in a certain way, we cannot be required to be motivated in a
certain way
This argument is far from persuasive. It rests on a dubious
understanding of the 'ought implies can' principle which has
very counter-intuitive consequences. If this understanding of the
Further arguments against this view can be found in The Foundations of Ethics, ell. 6.
'" Ibid. 120-1 and 118. '
principle were correct, then we would be freed from obligations
from which we do not think we are free. For example, suppose i
have promised John that I would pay back the money I owe him at
midday, and that John is counting on my doing so (he reaily needs
the money then). Suppose further that! have the money, and set oii
to meet him and pay him back. En route, however, 1 pass a casino
and, in a moment of weakness (and bad luck), lost: all of the money
at the blackjack table. If Ross's understanding of the 'ought implies
can* principle is correct, then as soon as 1 have lost the money 1 am
freed of the obligation to repay John. For although ! could have
chosen to pay him back, at noon before 1 lost the money, I can no
longer ciioose to do this, for it is not in my power to conjure up the
money at will. But this seems wrong. It seems that 1 still ought to
payback the money to John at midday even though I can no longer
choose to do so.
Ross might respond by saying that I, no longer have Ms obliga-
tion, but have acquired a different one—-namely, to apologize to
John and to find some other way of paving him back as ciuickly as
possible. This is partly right. I do acquire those extra obligations as
a. result of my action. But it is wrong to assume that these other
obligations replace the first one. I cannot so easily free myself of this
obligation. On the contrary, what i did was wrong because it meant
that I would no .longer he able to fulfil my obligation to John (to pay
him at midday). We can only make sense of" this if we assume that I
have made myself unable to fulfil an obligation that survives my
inability to fulfil it.
Ross's second argument against the view that we are required to
do morally good actions claims that this view leads to an infinite
regress. If we are required to act from a good motive, then we will,
at least sometimes, be required to act from the best motive. The
motive of duty is, Ross argues, the best motive (pp. 164—^). To act
from the best motive is, therefore, to act from a sense of duty. On
this view we will not have done what we should simply in doing
some act, say an act of helping another person. We will have doee
our duty only if we help someone from a sense of duty, for this is
what is required of us. But if helping someone from a sense of duty
is our duty, then, according to the theory under consideration, all of
this (helping from a sense of duty) roust: itself be done from a sense
INTRO D U CT1 ON xxxiii
of duty. This gives rise to a new duty (helping from a sense of duty,
from a sense of duty), which in turn must be done from a sense of
duty. This process will have to be repeated again, and again, leading
to a vicious infinite regress (pp. 5—6).
Ross regards this as a reductio adabsurdum of the view that we are
required to act from duty. This does not show that we cannot be
required to act from other good motives. But Ross rightly points out
that it would be paradoxical to assert that we are required to act from
good motives but can never be required to act: trorn the best (p. 6).
Since the most plausible monistic deontic theories are those that
attempt to ground the right in the good, if such theories are false it is
a safe bet that all monistic deontic th.eo.ries are false. If all monistic
theories are false, then if we are bound by any duties at all, pluralism
must be true.
Prima Facie Duties
This brings us to Ross's positive theory—his doctrine of prima
fade duties. This doctrine rests on. a distinction between 'prima
fade duly* and 'duty proper', or what Ross sometimes calls our
'actual duty* (pp. 19—20). It is tempting to understand this distinc-
tion as one between what merely seems to be our duty and what
really it our duty, but Ross is clear that it should not be understood
in this way. In fact the terns ''prima,facie duty* is doubly misleading,
for Ross intended it to pick out a feature of actions that is neither
a type of duty nor pritna fade (in the sense of being merely
The phrase 'prima fade duty* must be apologized for, since (i) it suggests
that what we are speaking of is a certain kinc! of duty, whereas it is in fact
not a duty, but something related in a special way to duty. Strictly speak-
ing, we want not a. phrase in which duty is qualified by an adjective, but a
separate noun, (z) 'Prima fade suggests that one is speaking only of an
appearance which, a moral situation presents at first sight, and which may
turn out to be illusory; whereas what I am speaking of is an objective fact
involved in the nature of the situation, or more strictly in an element of
its nature, though not, as duty proper does, arising from its whole nature,
(p. 20)
For an act to he prima facie right is not for us to have some kind of
duty to do this act, but is for this action to have some feature that
gives us a moral reason to do it. For some feature of an act to be
prima facie wrong Is not for us to have some kind of duty not to do it,
bur Is for it to have some feature that gives us a moral reason not to
do it. These features are not what we should do (our duty), but are
considerations which we (should) take into account in deciding
what it is we should do,
Ross does not make this clear in his official subjunctive condi-
tional definition of prima facie duties,39 or in his account of them in
terms of a certain tendency, or disposition.40 But this is, I believe,
what Ross means when he says that a prima fade duty 'is not a duty,
but something related in a special way to duty'. This interpretation
is given support by the tact that Ross discusses the doctrine of prima.
fade duties in a chapter called 'What Makes Right Acts Right?* in
The Right and the Good, and in 'Theories about the Ground of
Rightness* in The Foundations of.Ethics, So prima facie duties should
be understood as features that give us genuine (not merely appar-
ent) moral reason to do certain actions.
Because the term 'prima facie duty' is so misleading Ross was
keen to find some other term. 'Strictly speaking', he writes, 'we
want not a phrase in which duty is qualified by an adjective, but a
separate noun.* At the time of writing The Right ami, the Good., how-
ever, Ross could not think of a better term. He considered the term
"claim*, but rejects this for two reasons. First, it captures what he
wants to express from the wrong point of view—from the point of
view of the other person, the person who has a claim against me,
rather than of the agent. Secondly, it does not capture the fact that
we have a prima facie duty to ourselves, for only others can have a
claim against us (p. 20). In The Foundations of Ethics, however, he
accepts that the term 'responsibility* is a better way of expressing
what he was getting at with the terra, 'prima facie duty', Prima facie
duties are, he there states, responsibilities to ourselves and to others,
and what we should do (our duty proper) is determined by the bal-
ance of diese responsibilities.

'1 suggest 'pritna facie d u t y * , , , as a brief way of referring to the characteristic,,. which an
act has... of being an act which would be a doty proper if it were not at the same time of another
kind which is morally significant" (p, 19),
On p. 28 Ross characterizes the prima fade lightness of actions as their tending to be our
duty. See also Th& Foundations ofEAics, 86.
INTRO D U CT1 ON xxx v
He also accepts C. D. Broad's notion of fittingness, for unlike
lightness this allows of degrees. It makes no sense to say that a right
act is the one that is most right, but it makes perfect sense to say that
a right act is the one which is most fitting, or appropriate to the total
situation. Ross's account of duty proper in terms of fittingness is
very close indeed to the definition in terms of reasons offered above.
According to this definition, to say that an act is right is to say that it
has features that give us overall reason to do the action, whereas for
Ross it is to say that it has features that make it most appropriate to
the situation.
His account of tightness in terms of fittingness allows Ross to
understand prima facie right features of actions as those that are fit-
ting to some aspect of the situation. We can offer an account of prima
facie duties in terms of reasons also. To say that an act is prima facie
right is just to say that it lias a feature that gives us moral reason to
do the act (or, as Ross would put it, a feature that is fitting to some
aspect of the situation). So if one is troubled by the term 'fitting' one
can replace Ross's account in The Foundations of Ethics with art
account which replaces fittingness with reasons.
We have seen that. Ross rejects the view that all moral considera-
tions can he reduced to, or grounded in, one. as consequentiaiists
claim. Initially he lists seven f undarnental^>n/«a facie duties—-that is,
seven intrinsically (moral) reason-giving features of actions from
which all. other moral features derive their reason-giving force.
They are (i) considerations of fidelity, (2) of reparation, (3) of grat-
itude, (4) of justice, (f) of beneficence, (6) of self-improvement, and
(7) of non-maleficence (p. 21). Later on, however, he claims that (4)
to (6) can all be subsumed under the promotion of intrinsic value
(p. 27), for happiness, virtue, and justice are intrinsic goods. This
reduces the list of fundamental prima facie duties to five: fidelity,
reparation, gratitude, promotion of the good, and non-maleficence
(avoidance of the bad),
One might think that considerations of honesty should appear on
the list of basic moral considerations, but Ross thinks the reason-
giving nature of honesty is not basic but derivative. The fact that
our act is one of lying gives us moral reason not to do it only because
(and if) in lying we break an implicit promise to tell the truth, or
harm someone (pp. 54 5), In most eases one or both, of these basic
considerations will be relevant, but not always. For example, in a
game of Contraband"" one is expected to deceive, so there is no
implicit promise to tell the truth, no trust to be betrayed. If no one is
harmed by these lies, then lying fails under neither the prima. facie
duty ot fidelity nor that of non-maleficence. In such cases the fact
that our act is one of lying gives us no moral reason at all not to do it.
This shows that the reason-giving force of dishonesty is derivative,
rather than basic. The moral reason we have not to lie stems not
trom the very nature ot tying, but from considerations of fidelity
and non-maleficence. It is for this reason that when these basic con-
siderations do not apply, we have no moral reason not to lie.42
What this shows is that Ross is not simply listing whatever moral
obligations we think we have, something he is often accused of
doing. He is, rather, attempting to systematize as much as possible
the moral convictions we have. He does this by asking which moral
considerations are basic and which derivative, with the aim of find-
ing the considerations that underpin all the rest, but that are not
themselves underpinned by some other moral consideration. This
is basically the same enterprise consequentialists and Kantians are
engaged in. The only difference is that they think that all moral con-
siderations can ultimately be grounded in one, whereas Ross thinks
that they can be reduced to no fewer than five. This may not satisfy
everyone, but it is no more unsatisfactory than the fact that there is
an irreducible plurality of geometric axioms. The fact that there is a
plurality of axioms in geometry does not render it unsystematic.
Similarly, Ross's intuitionism is not unsystematic simply because he
thinks that there is an irreducible plurality of fundamental moral
Ross not only thinks that there is vertical structure (from basic to
derivative principles) but also some horizontal structure. For
although he thinks that any basic moral consideration can outweigh
any other, he does not think they carry equal normative weight.
He clearly thinks that certain basic considerations are weightier
than others. For example, he maintains that considerations of non-
maleficence, fidelity, and reparation are weightier than those ot
I borrow this example from Jonathan Dancy's Moral Reasons (Blackwelb Oxford, ^93),
Another example of a derivative/wTia facie duly is that ol: political obligation (pp. 27 8).
INTRO D U CT1 ON xxxvii
beneficence (pp. 2.2, 30). It is, however, bard fo specify what this
stringency amounts to in the abstract.43 it Is not that considerations
of non-maleficence, fidelity, and reparation always outweigh those
of beneficence. All that Ross can say is that considerations of, say,
fidelity can. be overridden only by a significant gain in well-being" for
others. But what is to count as significant will vary from case to case.
It is sometimes said that Ross cannot adequately account for
moral conflict.44 According to this view Ross's theory means that
what looks like a real conflict of duties turns out, after all, to be a
merely apparent conflict. For, it is said, in deciding which prima
facie duly generates our actual duty we are deciding which of the
conflicting duties is real and whicli merely apparent. On this view,
there turns out to be no real conflict at all, just an apparent one.
It should, however, be clear by now that this criticism rests on a
misunderstanding of the notion of zpritnafacie duty. Deciding what
our actual duty is is not, according to Ross, deciding which prima
facie duty is real and which merely apparent. It is, rather, deciding
which morally relevant consideration, gives us the weightier reason.
Deciding that the fact that we have made a promise gives us a
stronger reason to keep it than the fact that breaking it would pro-
duce more good gives us reason to break it is not deciding that the
latter fact gives us no moral reason at all. It is just deciding that we
have more reason to keep our promise than to break it. Ross shows
that he recognized this in claiming that when we decide that some
prima facie duty is overridden, we feel not guilt, but certainly com-
punction for die fact that we could not respond to the defeated duty
(p. 28).
In 1919 A. C, Ewing described Ross's doctrine of prima facie
duties as one of the most important discoveries in moral philosophy
in the twentieth century.4' This description maybe over the top, but
Ross's doctrine is certainly an important step in our understanding
of morality. One of the advantages of Ross's theory is that it can
accommodate the fact that we sometimes ought to break a promise
in order to prevent a bad outcome without conceding everything to
For Eoss's discussion of tlie stringency of i\Kprima fade duty to keep one's promises, see
The Foundations of£tkics^ IOQ^U
See e.g. J. Searfe, 'Prima Facie Obligations", in J. Raz (ed.), Practical Reasoning (Oxford
University Press: Oxford, 1978), 81-90.
Second Thoughts in Moral Philosophy (Rotitiedge & Kegan Paul: London, 1959)5 i2 ^-
the consequentialist (p. 18). Furthermore, before Ross, deontolo-
gists could only attempt to accommodate the pluralism that Is
deeply embedded in our moral thinking by Insisting that there is a
plurality of basic duties (duties proper, that is). The problem with
this, however, is that it turned ever? moral conflict into a tragic
moral dilemma whereby we art1 morally obligated to do two very
different actions in the same situation, but can only do one. In such
circumstances, whatever we do is wrong in the sense that we will fail
to do what we should no matter what we do. With his doctrine of
prima facie duties Ross frees pluralism from this difficulty. For this
doctrine allows us to think of moral conflict not as a conflict of
duties, but as a conflict of moral reasons, which, is, I believe, how
things are.
Pluralism about the Good
Discussions of Ross's pluralism tend to focus on his doctrine of
prima facie duties. But although Ross's views about the good are
almost never discussed, more than two-thirds of Tlte Right and the
GooJm devoted to this subject. Much of this is negative, attacking
those who deny that goodness is a simple, non-natural property. But
in Chapter V he outlines his positive theory about intrinsic goods-—
that is, about things that are good on their own account. Following
Moore, Ross divides such goods into those that have some parts that
have no value, and those that are either simple intrinsic goods, or
complex wholes all the parts of which are intrinsically good (p, 69).
He calls the latter 'ultimate goods*. Ultimately good things, unlike
other intrinsic goods, are good throughout—that is, contain noth-
ing that is not intrinsically good.
Although Ross followed Moore in distinguishing intrinsic from
ultimate goods, he differed from Moore in that he thought that for
the most part we can work out the intrinsic value of complex wholes
(which are not good throughout) from our knowledge of their ulti-
mately good parts.46 Consequently, he focuses on ultimate goods,
on the assumption that we can almost always work out which more
complex things are intrinsically good front our knowledge of which
" " . . . I regard the things thai are intrinsically but not ultimately good as owing, generally
speaking, their value to those elements in them that are ultimately good; as being good, in fact, "in
virtue of a part oi themselves", the other parts of them being irrelevant to their goodness' (p. 75).
INTRO D U CT1 ON xxxix
things are ultimately good. So he thinks that the evaluative has the
same foundationalist structure as the deontic. There is an irre-
ducible plurality of ultimate goods from which (almost) all other
intrinsic goods can be constructed, and in virtue of which they are
intrinsically good (p. 141).
Ross argues that there are three simple ultimate goods, and one
complex (which cannot be reduced to the value of its elements)
(p. 27).47 The simple goods are virtue, pleasure, and knowledge.
The complex good is the distribution of happiness in proportion to
merit (virtue). The first thing he claims is ultimately good is virtue.
He includes under virtue both virtuous dispositions and actions that
express these dispositions. The actions that express these disposi-
tions are actions clone from certain motives, such as the desire to do
one's duty, the desire to bring into being something good, or the
desire to give pleasure to others (p. 134), Hedonists would deny that
virtue is intrinsically good. They are monists about the good, and
hence insist that pleasure is the only intrinsic good. Ross argues that
we can see the falsity of hedonism if we consider a certain thought
experiment. If we consider two worlds that contain the same
amount of pleasure, though one is full of vicious and the other full
of virtuous people, we would not think that they are equally good.
Rather we would think that the one that has both, virtue and pleasure
would be the better world. This thought experiment shows the
iniplausibility of monistic hedonism (p. 134).
Ross uses the reverse thought experiment in support of the view
that pleasure is intrinsically good. If we imagine two worlds equal in
virtue, though in the first there is widespread and intense pleasure,
while in the second there is widespread and intense pain, we would
have no doubt that the first is die better (p. 135). This shows that we
do not think that virtue is the only thing that is good on its own
account, for the two worlds are equal in respect of virtue, though are
not equal in respect of'value,
Ross brings forward two other considerations in favour of the
view that pleasure is intrinsically good. The first is that we can only
explain our approval of kindness on the assumption that pleasure is
intrinsically good. If we did not think that pleasure is good, we
This view seems to commit Ross to Moore's doctrine of organic wholes, though in CIi, III
he is ratlier sceptical about this doctrine.
would not. think that the desire to give pleasure to others is a virtue.
Secondly, we think that virtue should be rewarded by happiness,
We would not, however, think this unless we thought that happiness
(here apparently equated with pleasure and the absence of pain) is
intrinsically good (p. 135).
The third (ultimate) intrinsic good is justice. By justice Ross
means the proportionment of happiness (again equated with pleas-
ure) to virtue, and misery to the vicious (p. 1.38),48 Unlike the first
two goods, this is a complex good, involving both pleasure and
virtue in a certain relation. It cannot, however, be reduced to the
value of virtue and pleasure separately. We can see this if we consid-
er two worlds with equal amounts of virtue, vice, pleasure, and pain,
hut which in one the virtuous are happy and the vicious miserable,
while in the other the virtuous are miserable and the vicious happy.
Although both worlds contain the same amount of virtue and pleas-
ure, they are not equally good. The former world is better than the
latter, To accommodate this we must, Ross thinks, allow justice as a
distinct and irreducible ultimate good. This is, he acknowledges,
one of the few clear instances of an organic whole. For both vice and
misery are intrinsically bad, but a whole involving the combination
of these elements is better than one in which the vicious are happy
(p. 72-)-
Finally Floss maintains that knowledge, and to a lesser degree,
true belief are intrinsically good (pp. 138—9). Once again, a compar-
ative thought experiment is used to make this clear.
... we ... may .. , help ourselves to realize the fact by supposing two
states of the universe equal in respect of virtue and of pleasure and of the
allocation of pleasure to the virtuous, but such that the persons in the one
had a far greater understanding of the nature and laws of the universe
than those in the other. Can any one doubt that the first would be a better
state of the universe? (p, 1.39)
Ross admits that knowledge of certain trivial facts, like the number
of storeys in a building, without knowledge of their relation to other
facts, may seem worthless. But he maintains that although such

Ross acknowledges that this is only one thing that we might mean by 'justice', bat it is this
that lie tliinks h ultimately good (p, 26).
knowledge has much less value than knowledge that involves
insight and understanding, it still has some intrinsic value
(P- >39)-
It is, however, not at all clear to me that knowledge, say, of the
number of grains of sane! on a certain beach, has any value whatso-
ever, Ross claims that even of such trivial facts right opinion is
better than wrong, and knowledge is better than right opinion. This
may be so, but does not show that such knowledge is good in itself.
One might think that one thing is better than another 'without
thinking that either is good, Ross is, however, on much stronger
ground with the view that knowledge and true belief that involve
insight and unde r. standing are good on their own account, independ-
ently of any benefit they might bring, (At least one would hope
this is true if one thinks that most philosophical enquiry has any
It may be objected that such knowledge has value only in so far as
it is a source of satisfaction to its possessor. But Ross rightly
responds to this; that one might, be equally satisfied with a false belief
that one thinks is true. But although we might he as satisfied with a
false belief as with genuine knowledge, we do not doubt that the tat-
ter state is better than the former (p. 140).
In. The Right and the Good, then, Ross maintains that there are
four ultimate goods: virtue, pleasure, justice, and knowledge. In
The Foundation.! of'Ethics, however, he rejected his earlier view
that pleasure is intrinsically good, owing to certain difficulties this
view entails,49 Ross lists a number of difficulties, the most signifi-
cant of which is that if pleasure were intrinsically good, then, we
would have a prima facie duty to produce pleasure not only for
others, but also for ourselves.*0 For if we have a prima fade duty to
promote the good, and our own pleasure is good, then we have a
prima facie duty to promote our own pleasure. It is not, however,
the case that we have a moral duty to promote our own pleasure. It
may be foolish not to promote our own. happiness when we have
the opportunity to do so, but it is not something that is morally
wrong, something about which we have reason, to feel guilty.
But although Ross abandoned his earlier view that pleasure is
The Foundations ofEtliics, 2ji B. *' Ibid. 271 2.
Intrinsically good, he never abandoned his pluralism about the
good (and the right)."

We saw earlier that Ross was keen to show that moral properties
could not be understood wholly in terms of non-moral properties.
This meant that he believed in the autonomy of ethics. To think that
ethics is an autonomous realm is to believe that ethics is a distinct
subject matter that cannot be subsumed under any other science.
We have also seen that Ross embraced a founclationalist epistemol-
ogy. He held, that we can know certain derivative moral truths only
because we know more basic truths. These more basic truths can in
turn be known only on the basis of our knowledge of even more
fundamental truths. But this cannot go on for ever, for we cannot
know an infinite number of truths. Following Aristotle,52 therefore,
Ross held that if we know anything derivatively, there must be
something that we know non-derivatively-—something that can be
known directly. We do seem to have derivative moral knowledge.
We know, for example, that lying is oftenprima facie wrong;, and
that we have a prima facie duty to obey the laws of our country.
According to Ross, we know that lying is prima facie wrong on the
basis of our knowledge that we have made an implicit promise to tell
the truth, or because lying will harm someone, and we know that
promise-breaking and harming others -AK prima jade wrong. But
how do we know that promise-breaking and harming others are
prima facie wrong? It does not seem that we derive this knowledge
from some other, more fundamental, moral proposition. If ethics
forms an autonomous realm we cannot derive this knowledge from
our knowledge of some non-moral premiss. We cannot derive our
knowledge that we should keep our promises from knowledge that
our society disapproves of promise-breaking, or that the institution
*' Akhough in The Foiuidations of Ethics Ross abandoned the view that pleasure Is intrinsical-
ly good, Ise still held that it is good in some sense (FQundauons., 275 ff,}.. But given that lie did not
abandon the view that we have %. prima fade duty to promote the good. It is unclear how he can
believe lhat pleasure is good in any sense and avoid the conclusion thai we have & prima facie duty
to promote our own pleasure.
Posterior Analytics, bk. *, chs. 2 and 3.
of promise-keeping has great utility. Since we cannot derive this
knowledge from our knowledge of other propositions (moral or
non-moral), then if we know that promise-breaking and harming
others hprima facie wrong we must know this directly.
This is true not only of the deontic, but of the evaluative also. We
know that certain things are instruraentally good, but this know-
ledge is derivative. It is based on our knowledge that certain things
are non-instrumentally good, and that something else is a means to
obtaining this good. But our knowledge that certain things are
intrinsically good does not seem to be derived from other evaluative
knowledge, and given the autonomy of ethics, this knowledge can-
not be derived from non-evaluative premisses, such as our know-
ledge that we desire or approve of that thing. Given that this
knowledge cannot be derived from moral or non-moral premisses,
if we know that certain things are intrinsically good, this must be
known directly—that is, non-inferenttally. If, therefore, we have
any moral knowledge at all—-be it cleontic or evaluative—we must
know certain fundamental moral propositions directly, on the basis
of an understanding of them. This means that, the most fundamental
moral propositions are self-evident.
But although Ross held that the principles of prima facie duty are
self-evident, he did not think that we know the stringency of these
duties. We only have a probable opinion about the degree of their
/jn'ffia/aaeobiigatoriness(p. y).5-'Furthermore, Ross was sceptical
about our ability to know what our actual duty is in particular situa-
tions. 'Our judgements about our actual duty in concrete situations
have none of the certainty that attaches to our recognition of the
general principles of [prima fade] duty 5 (p, 30). This is for three
reasons. First, we can never be sure that there is some prima facie
consideration that we have failed to notice. Second, 'any particular
act will in all probability in the course of time contribute to the
bringing about of good or of evil for many human beings, and thus
have a prima fade tightness or wrongness of which we know noth-
ing','4 And finally, even it we know which prima fade considera-
tions are relevant to the particular ease, we cannot know which is the
See also The Foundations of Ethics^ \ 88.
Ross seems to think of this second point simply as an illustration of'the first (p. 31). It is,
however, a distinct epistemological difficulty.
more stringent. So although Ross's view on moral epistemology
may seem overly optimistic in so far as he thinks that basic moral
propositions are self-evident, his overall view is in fact rattier pes-
Nonetheless, many moral philosophers think that Ross's epi-
steraologtcal intuitionism is utterly unacceptable, and reject it out of
hand. It is, however, worth asking what it is that is supposed to be so
untenable about this view. The most common response is that it pre-
supposes a mysterious sixth sense, a moral sense, by means of which
we apprehend the presence of moral properties-—intrinsic good--
ness a.nd prima fade Tightness. This objection is, however, based
upon a myth. If one looks at Ross's moral writings one will find no
reference to such a mysterious faculty. Is he committed to belief in
such a faculty? The simple answer is 'No'. Ross thinks that funda-
mental moral propositions are sell-evident. This means that we are
able to know these propositions solely on the basis of an under-
standing of them—that is, that they are knowable aprion?*1 If this is
right, then Ross is not committed to the view that we have a faculty
of apprehending moral properties, but only an ability to know cer-
tain propositions a priori, Mora] knowledge is no more mysterious
than any other a priori knowledge.
One may be suspicious of a priori knowledge in general on the
ground that it presupposes a mysterious faculty for apprehending
eternal truths. But I see no reason why we should think that a priori
knowledge requires such a mysterious faculty. All it requires is the
ability to understand and think. We should, therefore, quickly put
to one side the most common objection to epistemological intu-
itionism, and see if there are any objections with more substance to
It is sometimes argued that if the basic principles of duty were
self-evident, there would be no moral disagreement. There is moral
disagreement. Therefore, the basic principles of duty are not self-
evident. Ross considered this argument and rejected the conclusion.
^ It should be noted that, according to Ross, we come to know self-evident universal prin-
ciples by first knowing particular instances that fall tinder these principles. So, for example, we
come to know that promise-breaking \sprima fade wrong by first knowing that some particular
act of promise-breaking Is prima fade wrong. But knowledge of these particulars is, for Ross, sim-
ply a way of ascending to knowledge of the universal, knowledge which, once attained, can be
known directly (pp. 32—3).
* . . . the diversity of opinion on moral questions is found to rest not
on disagreement about fundamental moral principles, but partly on
differences in the circumstances of different societies, and partly on
different views which people hold, not on moral questions but on
Questions of fact'.*56 The sort of difference in circumstances Ross
had in mine! is illustrated by die presence or absence of a system for
the enforcement of public laws. Private vengeance may be right in
an unsettled state of society that lacks the proper provision for the
punishment of murderers, but wrong in a more settled society that
has such provisions. This moral difference is explained not by a dif-
ference in opinion about the fundamental principles of morality, but
by the fact that in an. unsettled state of society private vengeance
may be "the only way of'securing respect for life, while in a more set-
tled state of society this is better left to the arm of the law',57
Different moral views may also be explained by differences
about the non-moral facts.
.., the difference between those who think vaccination right and those
who think it wrong turns largely on a difference of opinion on the ques-
tion of fact whether vaccination does or does not prevent smallpox,
while both parties accept die principle that parents should try to save
their children from disease. And the difference of opinion between fox-
hunters and those who condemn fox-hunting turns largely on a differ-
ence of view as to the comparative intensity of the pain of the fox and the
enjoyment of his hunters.'8
One might not think that these examples accurately represent the
non-moral differences that underpin these moral disagreements.
Nonetheless, the point should be clear that many moral differences
rest on. different views about the non-moral (but morally relevant)
facts, rather than on differences about the fundamental principles
that Ross maintains are self-evident,
R.OSS acknowledges, however, that not all moral differences can
be accommodated in this way He writes:
While all men are probably at bottom agreed in thinking we ought
\primafade] to produce as much good as we can, and agreed also as to the
goodness of certain things virtue, intelligence, thought and happi-
ness—there is a real difference of opinion as to the comparative worth of
'" TheFourulationsofEthics, 18. " Ibid. ** Ibid.
different goods; and this is a difference not on an ordinary matter of fact
but on a. moral question."
This fact should, Ross thinks, weaken our confidence in out own
opinion. But we have already noted that he did not think that the
comparative stringency of our basic prima facie duties is self-
evident, even though their prima facie tightness is. One would
assume, therefore, that he does no t 'think that the comparative value
of certain goods is self-evident, even though it is self-evident that
certain things are intrinsically valuable. If this is right, then differ-
ence about comparative value is not damaging to his view about
self-evident, moral propositions.
Ross could have made the more general point that self-evident
propositions need not be obvious. Complex as well as simple math-
ematical proofs are self-evident in that we can know them apriority
reflection on their meaning. But in so tar as they are complex they
may not be at all obvious. Self-evidence, therefore, does not entail
obviousness. Once this is noted we can reject the first premiss of the
argument against self-evident moral principles—the claim that if
the basic principles of morality are self-evident there would be no
moral disagreement. This would, only be true if self-evident moral
principles must be obvious to everyone, as there is no general and
protracted disagreement about, obvious troths. But Ross is clear that
self-evident moral principles are not obvious, at least, not obvious
to everyone. He writes:
the nature of the self-evident is not to be evident to every mind however
undeveloped, but to be apprehended directly by minds which have
reached a certain degree of maturity, and for minds to reach the neces-
sary degree of maturity the development that takes place from genera-
tion to generation is as much needed as that which takes place from
infancy to adult life. (p. 12)
And later on he states that 'A prima facie duty is self-evident,
not in the sense that it is evident from the beginning of our lives, or as
soon as we attend to the proposition for the first time, but in the sense that
when we have reached sufficient, mental maturity and have givers suffi-

Ibid.j 59.
INTRO D U CT1 ON xl vii
cieni attention to the proposition it is evident without any need of proof,
or of evidence beyond itself, (p. 29)
To many ears talk of'mental maturity' and 'developed minds' may
sound arrogant, chauvinistic, or at best unacceptabiy complacent.
But Ross was neither arrogant nor complacent. Although he thinks
that there is moral progress, he does not think that we should accept
uncritically whatever moral system we inherit as the most devel-
oped and mature thinking on the issue. A sincere questioning in
ethics reflects, lie believes, 'a desire to get down to bedrock in
morality; and this is all to the good". He goes on to write:
Moral codes that will not survive such questioning do not deserve to sur-
vive it, and those that do deserve to survive it will do so, Magna est veri-
tas etpraevalet. Acquiescence in moral codes merely because they are
accepted by the society in which one lives spells death 10 progress in
moral insight.6"
These passages were not written by someone who thought that his
own thinking, or the moral thinking in his own culture, was so
developed and mature that it is beyond question.
Does the fact that one can have a good understanding of these
propositions yet deny that they are true show that: they cannot be
self-evident? I do not see how it could. For a proposition to be self-
evident is for it to be knowae/c on the basis of an understanding of
it. It does not follow from this that: if Pis self-evident and one under-
stands P, then one will believe that P. All it means is that one's
understanding of P provides a sufficient warrant for believing that
.P, not that one must recognise one *s understanding as a sufficient
Some object to the idea that certain moral propositions are
self-evident on the ground that this leads to dogmatism. Thus, for
example, Korsgaard claims that all that intuitionists can do in the
face of disbelief is insist that what they are saying is true, 61 and
Nowell-Smith writes: 'If I do not recognize the truth of your state-
ments of objective moral fact, you. can do nothing to convince me by
argument but are, like the subjectivist, thrown back on force or
«' Ibid. 21.
C. Korsgaard, The SourcesojNarmativity (Cambridge University Presw: Cambridge, 1996),
acquiescence'.*2 We should be on the guard against dogmatism,
and be aware that many of our moral beliefs may be die product
merely of upbringing, self-interest, or ideology. But a commitment
to self-evident moral propositions does not make intuitionists like
Ross especially vulnerable to dogmatism. For Ross does not dog-
matically assert the self-evidence of any moral proposition that
seems true to him, but claims this status only for those apparently
basic moral convictions that survive careful reflection,63 Such
reflection is aimed at getting a clear understanding of the proposi-
tion in question in order to ascertain whether we really do think it
expresses a morally basic truth. It also aims at providing arguments
for its truth and self-evidence. If the proposition survives such
reflection, then Ross can do more than merely assert the truth of the
moral proposition that someone else denies. He can take them
through the form of reflection that convinced him of the truth and
self-evidence of this proposition.
It may be thought that this is incoherent, that someone who
thinks that P is self-evident cannot offer an argument for P.M But
there is nothing about the fact that a proposition can be known sole-
ly on the basis of an understanding of it that rules out the possibility
that it can be known in some other way as well, i.e. that it is epi-
stemoiogically overdetermined."* It may not be that every self-
evident proposition is epistemologkally overdetermined, but the
thought that some are is in no way absurd.
In The Right and the Good Ross seemed to deny this. Here he
states that self-evident moral propositions 'cannot be proved, but
. . , just as certainly need no proof (p, 30). But elsewhere in The
Right and the Good-he makes only the more restricted claim that such
propositions do not need any proof,66 and despite the fact that he
sometimes makes the further claim that they cannot receive any
proof (justification) I do not think this further claim, expresses his

Ethics (Penguin: London, 19^6X46,
See ray 'Pleasure and Reflection in Ross's Ethical Iruuitiomsm', in Straiton-Lake (ect),
Ethical, fntuitionism: Re-Eyaluati&ns (Clarendon Press: Oxford, 2002).
Mary Warnock seems so think this (EAicssince tyao, 53),
^ See David O, Brink, 'Common Sense and First Principles in Sidgwick's Methods', Social
Philosophy and Policy, 11 (1994), i79--2O'i5 for an alternative view,
'" See e.g. p. 29.
considered view.*7 Three years before the publication of The Right
and the Goo J Ross explicitly states that 'the fact that something can
be inferred does not prove that it cannot be seen intuitively",*8 If he
thinks that some proposition can be inferred from (justified by)
other propositions and be self-evident, be clearly thinks that its
being self-evident does not rule out the possibility of a proof. So the
occasions in The Right and the Good where be denies this are, I think,
instances where Ross was riot as careful in what he said as he should
have been.
I have argued that the view that basic moral propositions are self-
evident is not as easily dismissed as it is so often thought to be. It
may, nonetheless, be that in the end 'Ross's moral epistemology can-
not be sustained. But because almost everyone is content with stan-
dard objections to epistetnological intuitionism-—objections which
are less than compelling—no one lias, to my knowledge, offered
decent arguments tor the rejection of this view. If Ross's epistetnol-
ogy commits us only to the view that basic moral propositions can
be known a priori, I for one, am happy to stick with it until such
arguments are put forward.

Most that is written on Ross focuses on his doctrine oiprima facie
duties, with passing reference to his metaphysics arid epistemology.
This is probably because this doctrine is regarded as separable from
his other views and hence need not be contaminated by their (per-
ceived) untenability. I have tried to provide a more rounded picture
of Ross's ethical intuitionism which situates his doctrine otprima
facie duties within a more general pluralism as well as a distinctive
metaphysics and epistemology.! have, however, left out a discus-
sion of Ross's views on a number of topics which, though interest-
ing, are not central to his intuitionism.! have said nothing about his
view about rights and their relation to obligations,69 his view on

In this I differ from Robert Audi's view in 'Imuitiomsm, Pluralism, and the Foundations of
Ethics', in W, SumotE-Armstrong and M. Thnmons (eds.), Moral Knowledge? (Oxford
University Press: New York, 1996).
"" 'The Basisof Ob jective Judgements in Ethics', International Journal of Ethics, 57(1927), 121.
" A.ff.lolGn.l\ofTheRightanittlu:Good.
punishment, 70 or his stance on the debate between subjective and
objective accounts ot duty (p. 43) (which changed dramatically in
The Foundations of Ethics11"), In general I have argued that many of
the standard criticisms of his views are at best less than compelling,
and are often based upon an accepted understanding that is little
more than a caricature, Ross was not content merely to list our com-
mon-sense moral opinions, and his metaphysics does not commit
him to a belief in otherworldly properties, or to the view that we
need a mysterious special faculty for perceiving these properties. I
think Ross should have abandoned his view that goodness and
rightness are simple properties, and suspect that he may have been
willing to do this if lie clearly distinguished the question of the sim-
plicity of these properties from the question of whether they are nat-
ural properties. But in the main I think his views are plausible and of
lasting significance.
"> App. » of Ch. U of 'flit Kigkanatlx GaoJ,
Ch, 'i. For an excellent discussion oi the arguments in Frkhard tbiH lcsd Floss to change Ms
mind, fees Jonathan L*)g,ncy^ 'Pnduutj, Duly, and Ignorance of Fact.*- in Sfrauon-Lake (ext),
Ethical IntuitiiW^m,


Asterisked cues in the text are to the Editor's Notes at the end of the
Books on Moral Philosophy
The Right and the Good ('Oxford University Press: Oxford, 1930;
Hackett: Indianapolis, 1988),
The Foundations of Ethics (Clarendon Press: Oxford, 1939).

Other Books
Aristotle (Methuen: London, 1923; Routledge: London, 1996),
Aristotle: Metaphysics (eel.), z vols,, (Clarendon Press: Oxford, 1924),
Aristotle: Physics (ed.) (Clarendon Press: Oxford. 1936).
Plato's Theory of Ideas (Clarendon Press: Oxford, 1951).
Kant's Ethical Theory: A Commentary on the 'Grundlegutig •{urMetuphysik
Jer Sitten "(Clarendon Press: Oxford, 1954).
Aristotle: ParvaNaturalia, (ed) (Clarendon Press: Oxford, 19^5).
Aristotle: Prior and Posterior Analytics (cd.) (Clarendon Press: Oxford,
Arixtotle: De Anima. (ed.) (Clarendon. Press: Oxford, 1961).

Article* on Moral Philosophy

'The Basis of Objective judgements in Ethics', International Journal of
Ethics, 37 (1927), 11.3—27.
'The Nature of Morally Good Action', Proceedings of the Aristotelian
Society, 29 (1928 9), 251 74,
"The Ethics of Punishment', Journal ojPhilosophical Studies, 4 (1929),

Other Writings
The Philosophical Aspect of the Theory of Relativity: A Symposium., by
A, S. Eddington, W. D. Ross, C, D. Broad, and F, A. Lindemann,
Mind, NS 29/116 (1920), 415—45.
"The Authenticity of Aristotle's Categories', in 'Comments and
Criticisms'y Journal of Philosophy, 36 (1939), 427-33.
'The Discovery of the Syllogism', Philosophical Review, 48 (1939),
'Some Thoughts on Aristotle's Logic", Proceedings of the Aristotelian
Society, 40 (1939—40), pp. i—xviii.
"The Date of Plato's Cnitjhis\ Revue Internationale de Pkilosophie, 9
(u)))), 187-96.

AL-Hiimi, A., 'Conditionality and Ross's Deontic Distinction',
Southwestern JournalofPhilosophy, 11 (1980), 7i>--88,
AQVIST, L., ^ Prima facie Obligations in Deontic Logic: A Chisholniian
Analysis Based on Normative Preference Structures', in C. Fehige
(eel,), Preferences (de Gruyter: Hawthorne, Nev., 1998).
ATWKLL, j., * Ross and Prima Fade Duties', Ethics, 88 (1978), 240-9,
BAUMRIN, B, H., * Prima Facie Duties', Journal oj Philosophy, 62 (1965),
736 8.
BECK, L. W,, 'Sir David Ross on Duty and Purpose in Kant*, Philosophy
and Phenomenological Research, 16 (1955), 98—107.
BERAN, H., 'Ought, Obligation and Duty', Australasian Journal at
Philosophy, 50 (1972), 207 21.
BHENNAN, S. K,, "Ross, Promises and die Intrinsic, Value of Acts',
Lyceum, i (1989). 43—56,
BROWNE, S. S, S-, 'Right Acts and Moral Actions', Journal of Philosophy,
42(1945X505 15,
CHISHOI.M, R. M., 'Kpistemic Reasoning*, Dialeaica, 36(1982), 169—77.
DANCY, j.. "Ethical Particularism and. Mttrally Relevant: Properties',
Mind, 92 (1983), 530- ,47,
*An Ethic of Prima fade Duties', in P. Singer (ed.), A Companion
to Ethics (Blackwell: Oxford, 1991),
'"Wiggins and Ross*, Utilitas, 10 (1998), 281—5.
DARWALt, S., "Under Moore's Spell', Utsluas, 10(1998), 286 i)i.
EDWARDS, R. B., *A Criticism of Ross's Hypothetical "I Can" \Mind, 69
(i 960),'80-3.'
Fox, A. C,, 'The Right and the Good Once More *, Australasian Journal
of Philosophy, 28 (1950), 1—28,
FRAZIER, R., 'Moral Relevance and Cetera Parihus Principles', Ratio, 8
(1995), "3-27-_
GALVIN, R. F.,']Prima Facie Duty; An Analysis', Southwest Philosophical
Studies, 1.00/2 (1990), 41-52.
GARCIA, J., "On High-Mindedness , Proceeding.1; of the American Catholic
Philosophical Association, 6-5 (1990), 98—107.
'The Right and the Good', Philosophia, 21 (1992), 235 6,
GAY, R., "Ethical Pluralism: A Reply to Jonathan Dancy', Mind, 94
(1985), 250-62,
( . 1 1 , 1 1 1 , ' ! . I!. 'On f <n. !">> 'visibility of Duties', jour rial, of i/aliie fmjuiry,
i . VH> >}. lu- 2 * }
HovihiH H M l o ' - S n ! , Pluralism versus Rule-Consequentialisrn',
I//?*/ i " { O o 1 ^ <| ji - i ^
Tnv|'!\ iu'Mf titon-I Ac\Mind, 106(1997),759—60,
] M «• H i f , , "\1<-H (HI !''i>'ia Fade Duties \ Journal aj'Philosophy, 6}
( P)H<>), ,?1 }
' N ' i ' o ' " » ' )i ikif» ibopi jPwna.Fa<rie'.Duties',Philosophy^46(1971),
J<K,> i.
"I''p'', ain 1 Boss's Theory of Prima Facie Duties',
l)l->lui;'t<>, ! . (^li) -M !, (; - >)7,
|i)N r V ]', J)r>utii l ahunt / i//?w /*««'<." Duties', Philosophy^ 45 (1970),
39 54-
LEWIS, H, 'Moral Freedom in Recent Ethics*, in W. Sellars and
J, Hospers (eds,), Readings in Ethical Theory (Appieton-Century-
Crofts: New York, 1952).
MACBEATH, A., 'Duty', Philosophy, 23 (5948), 99 ? t ) ,
McCi.osK.KY, H. j., 'Ross and the Concept of a Prima Facie Duty',
Australasian Journal of Philosophy, 41 (1963), 336—45,
McCoNNEUL, T,, "Ross on Duty and Ignorance*. History of Philosophy
Quarterly, 5 (1988), 79 95.
MCNAUGHTON, D., 'An Unconnected Heap of .Duties?', Philosophical
Quarterly^ 46 (1996), 435—47.
MORREAU, M,, 'Prima Facie and Seeming Duties', Siiulia Logica, 57
(1996), 47-71.
NEBLETT, W, R,, 'W, D, Ross and the Nature of Moral Obligation*,
Journal'oj Value Inquiry, j (1973), 40—51,
NYSTEOT, H., "The Problem of Duty and Knowledge', Philosophy, 26
(1951), 333~4<S.
PIETROSKI, P., 'Prima Facie Obligations: Ceteris Paribus Laws in Moral
Theory \Ethics, 103 (1991), 489—515,
ROSEN, B., *!n Defense of W, D, Ross*. Ethics,. 78 (1968), 237-41.
SAY WARD, C., 'W. D. Eoss on Acting from Motives', Journal of Value
Inquiry, 22 (1988), 299—306,
SEARLE, J, R., 'Prima Fade Obligations', in J. Raz (eel). Practical
Reasoning (Oxford University Press: Oxford, 1978).
SHAW, L. M,, "Can we Choose our Motives?", Australasian Journal of
Philosophy, 22 (1944), 93—105,
SHOPE, R. K,,'Prima Facie Duty \Joumctl ofP/iiiosop/iy, 62 (1965), 279 86.
SNARE, R, "The Definition of Prima Fade Duties', Philosophical
Quarterly, 24 (1974), 235-44.
STRASSER, ML, 'G uilt. Regret and Prima Facie Duties', Southern journal
of Philosophy, 25 (1987), 133—46,
'Ross on the Sense of Duty*, Philosophical Paper.*, 16 (1987),
STRATTON-LAKE, P., 'Can Hooker's Rule-Coosecjuentialist Principle
Justify R,oss's Pr/ffia Facie Duties?', Mind, 106 (1997), 751 --8.
WHEELER, A. M., ^Prima Facie and Actual Duly', Analysis^ 37 (1977),
WIGGINS, D., 'The Right and rhe Good said W. D, Ross's Criticism of
Consequentialtsm', IJtilitas, to (1998), 261-80; and in A. O'Hear
(ed,), Philosophy, the Goad., the True and the Beautiful, Philosophy, 47,
suppl. (2000), 175-95.
WOLF, J. C., 'Ein Piuralisraus von Prinui Facie Pflichten als Alternative
zu monistischen Theorien der Ethik', Zeitsckrift fur phUasophuche
Forxchung, 50 (1996), 601—10.


ATKINSON, G. M., 'A Defense of Intuitions: 'Their Roie in Moral
Knowledge*, Proceedings oj ike .American Catholic Philosophical
Association, 64 (1990), 107—17,
AUDI, R., "Ethical ileflectionism.', Monist, 76 (1991), 295—515.
'Intukionism, Pluralism, and the Foundations of Ethics', in
W, Sinnott-Arrnstrong and M, Timmons (eds.), Moral Knowledge?
(Oxford University Press: New York, 1996).
'Moderate Intuitionism and the Epistemology of Moral
Judgment', jEt/fc'ca/ Theory and Moral Practize, t (1998), 15—44.
'Self-Evidence', Philosophical Perspectives, 13 (1999), 205--28,
*A Kantian Intuitionism', A/zW, no (2001), 601 35.
B.ALGUY, J., The Foundation of Moral Goodness (1728), facs. edn,
(Garland: New York, 1978), and in L. A. Selby-Bigge (ed.), The
British Moralists: Being Selections from the Writers Principally oj the
Eighteenth Century, ii (Dover; New York, 1965).
BINKLEY, L. J,, Contemporary Ethical Theories (Philosophical Library:
New York, 1961).
BRINK, D. O,, 'Common Sense and First Principles in Sidgwick's
Methods, Social Philosophy and Policy, ll (1994), 1.79—20!.
BROAD, C. D., Five Types of Ethical Theory (Kegan Paul: London,
BRODY, B, A., "Intuitions and Objective Moral Knowledge*, Monist, 62
(1979), 44<>-1<5.
BURTON, R. G., 'Neointuitionism: The Neglected Moral Realism',
Southernjournal of Philosophy, 25 (19877), 147—12,
B'UTLEK, J., Fifteen Sermnns Preached at the Rolls Chapel (\j2,6}y in The
Works oj Bu/iop Butler^ ed, j. H. Bernard, i (Macniillan: London,
1900); and in L. A, Selby-Bigge (ed.), The British Moralists; lieintf
Selections from the Writers Principally of the Eighteenth Century^ i
(Dover; New York, 1965),
CARROT, E. F,, The Theory of Morals (Oxford University Press:
London, 1930).
Ethical and Political Thinking (Clarendon Press; Oxford, 1947).
CLARICE, S., A Discourse Concerning the 'Unalterable Obligations of
Natural Religion (1706), in The Works of Samuel Clarke, ii (J. & P.
Knapton: London, 1738), lacs. edn. (Garland: New "York, 1978); and
in L. A.Selby-B igge (eci), Tlie British .Moralists: Being Selections from,
the. Writers Principally of the Eighteenth Century,. ii (Dover; New York,
COHEN, C., 'Natural and Non-Natural Queues',Journal of Philosophy,
CUDWORTH, A Treatise Concerning Eternal and Immutable Morality
(1731), in The Collected Works of Ralph. Cudwonh, ii (Olms:
Hildesheim, 1979), and in L, A. Selby-Bigge (ed.). The British
MoTcdwts; fieing Selections from the Writers Principally oj the
Eighteenth- Century, ii (Dover: New York, 1965),
DALY, C. B., Inter-War British Ethics, 2: Critics of Intuitioriism",
Philosophical Studies (Ireland), 15 (1965), 96—111.
DANCY, J,, 'Intuitionism in Meta-Epistemology', Philosophical Studies,
42 (1982), 395-408.
* Intuitionism', in P. Singer (ed,), A Companion to Ethics (Blackwell:
Oxford, 1991).
DARWALL, S., "Moore to Stevenson', in R. J, Cavalier (eds.), Ethics in
the History of Western Philosophy (5t Martin's Press: New York,
The British Moralists and the Internal 'Ought"': 1640—1340
(Cambridge University Press: Cambridge, 1995).
DURFEE, H. A,, "The A Priori, Intuitionism and Moral Knowledge',
Philosophical Studies (Ireland), 27 (1980), 51-66.
EWING, A. C., The Definition of Good (Routledge & Kegan Paul:
London. 1947).
Ethics (Macmillan: London, 1947).
The Fundamental Questions of Philosophy (Routledge & Kegan
Paul: London, 1951).
KwiNG,A. C,, Second Ikoughts in Mor<dPliUosophy(&o\)A\edQK&K£<gjl.n
Paul: London, 1959).
FRAZJER, R. L., 'Intuitionistn in Ethics*, in E. Craig (ed.), Ruuiledge
.Encyclopedia of Philosophy, vl (Routlecige: London, 1998),
G Atrr, B., 'Moral Pluralism', Philosophical Papers, 22 ('99'?), 17-40.
'Rag-Bags, Disputes and Moral Pluralism*, Utilitas, n (1999),
GEIGER, G., "Why there are No Objective Values; A Critique of Ethical
Intiiitionism from an Evolutionary Point of View*, Biology and
Philosophy, 7 (1992), 315—30.
HKYTING, A., Irituitusnum: An Introduction (North-Holland:
Amsterdam, 1956).
HILL, T. E., Contemporary Ethical Theories (Macmillan; New York,
HUDSON, W. ~D,, Ethical Inflationism (Macraillan: London, 1967),
A Century of Moral P/tilos0j)/iy(St Martin's Press; New York, 1980).
ModernMoralPhilosophy, 2nd edn. (St. Martin's Press: New York,
JOHNSON, O, A,, 'Highrness, Moral Obligation, and Goodness', Journal
of Philosophy, 50 (1953), 597-^07-
'Ethical Intiiitionism: A Restatement', Philosophical Quarterly, 7
(1957), (93-203,
Riglitness and Goodness: A Study in Contemporary Ethical Theory
(Martinus Nijhoff: The Hague, 1959).
JOSEPH, H. W. B, Some Problems in Ethics (Clarendon Press: Oxford,
KARANI, R. N., 'Vindication of Ethical Intiiitionism', Mind, 71 (1962.),
KEKES,J.,'Moral Intuition , American Philosophical Quarterly,, 23 (1986),
»3 93-
LEVENBOOK, B. B., and GIRILL, T, R., 'The Parti-Resultantness of
Requirement: An Explanation that Failed', Philosophical Studies, 37
(1980), 237-49-
LUCAS, ], R,, 'Ethical Intuitionism II', Philosophy, 46 (1971), i—10.
McCLOSKEY, H., "Intuitionism and Rational Ends: Regan's Account of
the Intuitionist's Dilemma'Journal of Faltte Inquiry, 9 (1975), 59—66,
MAC!NTYB.E, A, C., A Short History of Ethics; A History of .Moral
Philosophy from the Homeric Age to the Twentieth Century (Macmillan:
New York, 1966); and edn, (London: Routledge, 1998).
MACLAGAN, W. G., 'On being Sure of One's Duty*, Pliilosophical
Quarterly, i (1950), 35—49.
McNAUGHTON, D., 'Intuitiotiism', in H. LaFollette (ed,), The Blackwell
Guide to Ethical Theory (Blackwell: Oxford, 2000).
MARGOUS, J., *G. E. Moore and Intuitionism', Ethia, 87 (1976), 35- -48.
MOORK, G, E., Principia Ethica (Cambridge University Press:
Cambridge, 1903; rev, ecln. 199}).
Ethics (Oxford University Press: London, 8966).
"f he Conception of Intrinsic Value', in Principia Ethica, rev. edn,
(Cambridge University Press: Cambridge, 1993).
MUIRHEAD, J. H., 'The New Deontology', Ethics, 50 (1940), 441-9,
NELSON, M., 'inruitionism and Conservatism*, Meiaphilosophy, 21.
(1990), 182-93.
'Intuitionism and Subjectivism', Metaphilosophy, 22 (1991),
j 15 21.
"Morally Serious Critics of Moral Intuitions', Ratio, 12 (1999),
NOWELL-SMITH. P. H., Eifiics (Penguin: London, 1956).
ODEGARB, D,. "On Two Arguments against Moral Certainty*, Mind, 90
PHII.LIPS, R. L., 'Intuitionism Rf-visited*,,Journal of Value Inquiry, 6
(1972), 185 99.
PRICE. R.. A Review of the Principal Questions in Morals (1758), ed. D, D.
Raphael (Clarendon Press: Oxford, 1974); and in L, A. Selhy-Bigge
(ed.), The British Moralists." .Being Selections from the Writers
Principally of the Eighteenth Century n (Dover: New York, 196)).
PRICHAHD, H. A., Moral Obligation, eel. W. D. Ross (Clarendon Press:
Oxford, 8949).
RAPHAEL, D, D,, 'Sklgwick on Intuitionism', Monist, 58 (i 974), 405—19.
Moral, Philosophy (Oxford University Press: Oxford,, 1981).
'LesIntuitionnistesd'Oxtbrd',^rcAfV«jrdfePA//ar<yAifi, 57 (1994),
RASHDALL, H.,A Theory of Good and Evil (Clarendon Press: Oxford,
REGAN, T., 'McCloskey on Rational Ends: The Dilemma of
ln\xaiioi\Km,\Journal of Falue Inquiry, 9 (1975), 42—58.
REID, T., Essays on the Active Powers of Man, (1788), in The Works of
ThomasReid(i%6$), is, ed. W. Hamilton (Thoemmes: Bristol, 1994);
and in D. D. Raphael (ed.), British.Moralists t.65o—>.8oa, ii (Clarendon
Press: Oxford, 1974),
SCHNEEWIND, j. B., 'First Principles and Common Sense Morality in
Sidgwick's Ethics', Archiv fiir Gesch.ich.te cler Philosophic, 45 (1961),
SCHN'EEWIND, ]. B.. Sidgwick's Ethics and Victorian Moral Philosophy
(Clarendon Press; Oxford, 1977),
SHAW, W. H., "lutuitionism and Moral Philosophy*, American
Philosophical Quarterly, fj (1980), 127—34.
SIDGWICK, H,, The Methods of Ethics, 7th edn. (Macmillan: London,
1907 and 1967); and in The Works of Henry Sidgwick, ed. J. Slater,
(Thoemmes: Bristol, 1996).
Outlines of"the History of Ethics for English. .Readers (Macmillan;
London, 1954); and in The Works of Henry Sidgwick, ed. J, Slater, iv
(Thoemmes: Bristol, 1996),
SIMON, €", J., "On Defending a Moral Synthetic A Priori', Southern
Journal of Philosophy, 26 (1988), 217-33.
SINNOTT-ARMSTRONG, W, 'Intuitionism', in L, C. Becker (eel), The
Encyclopedia of Ethies, I (Garland: New York, 1992).
SMITH, M, B. E., 'Ethical Tntuitionism and Naturalism; A
Reconciliation', Canadian Journal of "Philosophy, 9 (1979), 609—29.
STRATTON-LAKE, P., "Why Externalism is not a Problem for Ethical
Intuitionists', Proceedings of the Aristotelian Society, 99 (1999)- 77"~9°-
Ethical Intuitionism; Re-Evaluations (Oxford University Press:
Oxford, 2002),
STRAWSON, P. F., "Ethical Intuitionism', Philosophy, 24 (1949), 23 33.
SWANTON, C., "The Rationality of Ethical Inruittonism', Australasian
Journal of Philosophy, 6f (1987-},, 172—81,
URMSON, J. O., ' A Defence of Intuitionism', Proceedings oj the
Aristotelian Society. 75 (1975), 111—19.
WARNOCK, G. J., Contemporary Moral Philosophy (St Martin's Press;
New York, 1967).
WARNOCK, M., Ethics since 1,900 (Oxford University Press: London,
WHITTEMORE, R, C., 'Does the Neo-Intuitionisi Theory of Obligation
Rest on a Mistake?*, Tulane Studies in Philosophy, 6 (1957), 101—27.
WILLIAMS, B,, 'What does Intuitionism Imply?', in J. Dancy (ed.),
Hitman Agency (Stanford University Press; Stanford, 1988); and in
Williams, Making Sense of Humanity and Other Philosophical Papers
1982—1993 (Cambridge University Press: Cambridge, 1995).

T HE purpose of this inquiry is to examine the nature, rela-

tions, and implications of three conceptions which appear
to be fundamental in ethics—those of'right*, 'good' in general,
and 'morally good*. The inquiry will have much in common
with the inquiries, of which there have been many in recent
years, into the nature of value, and I shall have occasion to
discuss some of the more important theories of value; but my
object is a more limited one, I offer no discussion, except at most
a purely incidental and illustrative one, of certain forms of value,
such as economic value and beauty. My interest will through-
out be ethical, and value will be discussed only so far as it seems
to be relevant to this interest,
I propose to begin with the term 'right*. A considerable
ambiguity attaches to any attempt to discuss the meaning of
any term. Professor G. E, Moore has well indicated three main
objects that such an attempt at definition may have. 'When we
say, as Webster says, "The definition of horse is, 'A. hoofed
quadruped of the genus Equus*," we may, in fact, mean three
different things, (i) Wemaymean merely: "When I say'horse',
you are to understand that I am talking about a hoofed quad-
ruped of the genus Equus." This might be called the arbitrary
verbal definition.... (2) We may mean, as Webster ought to
mean: "When most English people say 'horse', they mean a
hoofed quadruped of the genus Eqiius." This may be called
the verbal definition proper. , . . But (3) we may, when we
define horse, mean something much more important. We may
mean that a certain object, which we all of us know, is composed
in a certain manner: that it has four legs, a head, a heart, a liver,
etc., etc., all of them arranged in definite relations to one
We must ask ourselves whether, in discussing the meaning
of 'right*, we are attempting any one of these kinds of definition,
or something different from them all, I certainly do not wish
1 Principw JEtfaica, 8*
merely to indicate a sense in which I propose to use the term
"right*. I wish to keep in touch with the general usage of die
word. While other things may be called 'right* (as in the
phrases 'the right road', 'the right solution*), the word is specially
applied to acts, and it is the sense (by general consent a very
important one) in which it is so applied that I wish to discuss.
But we must be prepared to find that the general usage of the
word is not entirely consistent with itself. Most of the words
in any language have a certain amount of ambiguity; and there
is special danger of ambiguity in the case of a word like 'right',
which does not stand for anything we can point out to one
another or apprehend by one of the senses. Even with words
that do stand for such things there is this danger. Even if two
people find that the things the one calls red are just the things
the other calls red, it is by no means certain that they mean the
same quality. There is only a general presumption that since
the structure of their eyes (if neither is colour-blind) is pretty
much the same, the same object acting on the eyes of the two
men produces pretty much the same kind of sensation. And in
the case of a term like 'right', there is nothing parallel to the
highly similar organization of different people's eyes, to create
a presumption that when they call the same act right, they mean
to refer to the same quality of it. In point of fact, there is a
serious difference of view as to the application of the term
'right'. Suppose, for instance, that a man pays a particular debt
simply from fear of the legal consequences of not doing so,
some people would say he had done what was right, and others
would deny this: they would say that no moral value attaches
to such an act, and that since 'right* is meant to imply moral
value, the act cannot be right. They might generalize and say
that no act is right unless it is done from a sense of duty, or if
they shrank from so rigorous a doctrine, they might at least
say that no act is right unless done (tomsome good motive, such
as either sense of duty or benevolence.
This difference of view may be due to either of two causes.
Both parties may be using 'right* in the same sense, the sense
of'morally obligatory*, and differing as to the further character
an act must have in order to have this quality. Or the first
party may be using 'right* in this sense, and the second in the
sense of 'morally good'. It is not clear to me which of these
two things is usually happening when this difference of view
arises. But it seems probable that both things really happen—
that some people fail to notice the distinction between "right*
and 'morally good', and that others, while distinguishing the
meaning of these terms, think that only what is morally good
is right. A discussion of the first of these positions only is
strictly in point here, where we are discussing the meaning of
'right'. It seems to me clear that 'right' does not mean the same
as 'morally good*; and we can test this by trying to substitute
one for the other. If they meant the same thing we should be
able to substitute, for instance, 'he is a right man' for 'he is a
morally good man'j nor is our inability to do this merely a
matter of English idiom, for if we turn to the sort of moral
judgement in which we do use the word 'right*, such as 'this is
the right act*, it is clear that by this we mean 'this act is the act
that ought to te done\ 'this act is morally obligatory'; and to
substitute either of these phrases for 'morally good* in 'he is a
morally good man* would obviously be not merely unidio-
matic, but absurd. It should be obvious, then, that 'right* and
'morally good* mean different things. But some one might say
that while 'morally good' has a wider application than 'right*,
In that it can be applied to agents as well as to acts, yet when
applied to acts they mean the same thing. I should like there-
fore to convince him that 'right act' cannot mean the same as
'act that ought to be done* and also the same as 'morally good
act'. If I can convince him of this, I think he will see the pro-
priety of not using 'right act' in the sense of 'morally good act*,

But we ought first to note a minor difference between the

meaning of 'right* and the meaning of 'something that ought
to be done' or 'that is my duty* or 'that is incumbent on me'.
It may sometimes happen that there is a set of two or more acts
one or other of which ought to be done by me rather than any
act not belonging to this set. In such a case any act of this set
is right, but none is my duty; my duty is to do "one or other*
of them. Thus 'right' has a somewhat wider possible applica-
lion than 'something that ought to be done* or any of its equi-
valents. But we want an adjective to express the same meaning
as 'something that ought to be done", and though we have
'obligatory' at our disposal, that also has its ambiguity, since
it sometimes means 'compulsory'. We should have to say
"morally obligatory* to make our meaning quite clear; and to
obviate the necessity of using this rather cumbrous expression,
I will use 'right* in this sense. I hope that this paragraph will
prevent any confusion arising from this slightly inaccurate
Some might deny the correctness of the distinction just
drawn. They might say that when there are two or more acts
one or other of which, as we say, we ought to do (it not being
our duty to do one rather than another), the truth is that these
are simply alternative ways of producing a single result, and
that our duty is, strictly, not to do 'one or other' of the acts,
but to produce the result; this alone is our duty, and this alone
is right. This answer does, I think, fairly apply to many cases
in which it is the production of a certain result that we think
obligatory, the means being optional: e.g. to a case in which it
is our duty to convey information to some one, but morally
immaterial whether we do so orally or in writing. But in
principle, at any rate, there may be other cases in which it is
our duty to produce one or other of two or more different
states of affairs, without its being our duty to produce one of
them rather than another; in such a case each of these acts will
be right, and none will be our duty,

If it can be shown that nothing that ought to be done is ever

morally good, it will be clear a fortiori that 'morally good* does
not mean the same as 'that ought to be done'. Now it is, I
think, quite clear that the only acts that are morally good are
those that proceed from a good motive; this is maintained by
those whom 1 am now trying to convince, and I entirely agree.
If, then, we can show that action from a good motive is never
morally obligatory, we shall have established that what is
morally good is never right, and a fortiori that 'right* does not
mean, the same as 'morally good', That action trom a good
motive is never morally obligatory follows (i) from the Kantian
principle, which is generally admitted, that 1 ought* implies
*I can*. * It is not the case that I can by choice produce a certain
motive (whether this be an ordinary desire or the sense of
obligation) in myself at a moment's notice, still less that I can
at a moment's notice make it effective in stimulating me to act.
I can act from a certain motive only if I have the motive; if not,
the most I can do is to cultivate it by suitably directing my
attention or by acting in certain appropriate ways so that on
some future occasion it will be present in me, and I shall be able
to act from it. My present duty, therefore, cannot be to act here
and now from it,
(2) A similar conclusion may be reached by a reductio ad
absurdum. Those who hold that our duty is to act from a cer-
tain motive usually (Kant is the great exemplar) hold that the
motive from which we ought to act is the sense of duty. Now
if the sense of duty is to be my motive for doing a certain act,
it must be the sense that it is my duty to do that act. If, there-
fore, we say 'it'is my duty to do act A from the sense of duty*,
this means 'it is my duty to do act A from the sense that it is my
duty to do act A'. And here the whole expression is in contra-
diction with a part of itself. The whole sentence' says 'it is
my duty to-do-act-^-from-the-sense-that-it-is-my-duty-to-do-
act-v^*. But the latter part of the sentence implies that what I
think is that it is my duty to-do-act-^ simply. And if, as the
theory in question requires, we try to amend the latter part of
the expression to bring it into accord with the whole expression,
we get the result 'it is my duty to do act A from the sense that
it is my duty to do act A from the sense that it is my duty to do
act A', where again the last part of the expression is in conflict
with the theory, and with the sentence as a whole. It is clear
that a further similar amendment, and a further, and in the end
an infinite series of amendments would be necessary in the
attempt to bring the last part of the expression into accordance
with the theory, and that even then we should not have suc-
ceeded in doing so,
Again, suppose that I say to you 'it is your duty to do act A
from the sense of duty'; that means 'it is your duty to do act A
from the sense that it is your duty to do act A*, Then / think
that it is your duty to act from a certain motive, but I suggest
that you should act under the supposition that it is your duty
to do a certain thing, irrespective of motive, I. e, under a sup-
position which I must think false since it contradicts ray own.
The only conclusion that can be drawn is that our duty is to
do certain things, not to do them from the sense of duty,1
The latter of these two arguments ( (i) and (2) ) cannot be
used against those who hold that it is our duty to act from some
other motive than the sense of duty; the sense of duty is the
only motive that leads to the infinite series in question. But
the first of the two arguments seems in itself sufficient against
any theory which holds that motive of any kind is included in
the content of duty. And though the second argument does not
refute the view that we ought to act from some other motive,
it would be paradoxical to hold that we ought to act from some
other motive but never ought to act from a sense of duty,
which is the highest motive.2
Let us now return to the three senses in which Professor
Moore points out that we may understand an, attempt to define
a certain term,3 So far, the position we have taken up with
regard to 'right* includes something of each of the first two
attitudes he distinguishes. In using "right* as synonymous
(but for the minor distinction already pointed out) 4 with 'what
is my duty', and as distinct from 'morally good*, I believe I am
conforming to what most men (if not all men) usually mean
when they use the word. But I could not maintain that they
always use the word in this way. I am, therefore, to some
extent adopting the first of the attitudes he distinguishes, and
expressing my own intention to use 'right' in this sense only.
And this is justified by the great confusion that has been intro-
duced into ethics by the phrase *a right action* being used
sometimes of the initiation of a certain change in the state of
It should be added, however, that one, and an important one, of our duties is to
cultivate in ourselves the sense of duty. But then this is the duty of cultivating in
ourselves the sense of duty, and not of cultivating in ourselves, from the sense of duty,
the sense of duty,
* If any one doubts that it is, I beg him to refer to pp. 164-5, where I give reasons in
support of the contention,
J Cf. p. i. 4 pp. 3-4.
affairs irrespective of motive, and at other times of such initia-
tion from some particular motive, such as sense of duty or
benevolence. I would further suggest that additional clearness
would be gained if we used 'act' of the thing done, the initiation
of change, and "action* of the doing of it, the initiating of change,
from a certain motive. We should then talk of a right act but
not of a right action, of a morally good action but not of a
morally good act. And it may be added that the doing of
a right act may be a morally bad action, and that the doing
of a wrong act may be a morally good action; for 'right' and
•'wrong* refer entirely to the • thing done,* 'morally good* and
'morally bad' entirely to the motive from which it is done, *A firm
grasp of this distinction will do much to remove some of the
perplexities of our moral thought.*
The question remains, what attitude we are to take up to-
wards Professor Moore's third sense of 'definition'. Are we to
hold that 'right* can be defined in the sense of being reduced
to elements simpler than itself? At first sight it might appear
that egoism and utilitarianism are attempts to define 'right'—
to define it as "productive of the greatest possible pleasure to the
agent' or as 'productive of the greatest possible pleasure to
mankind'| and I think these theories have often been so under-
stood by some of those who accept them. But the leaders of
the school are not unanimous in so understanding their theory.
Bentham seems to understand it so. He says * that "when thus
interpreted* (i.e. as meaning 'conformable to the principle of
utility'), 'the words ought and right,,. and others of that stamp,
have a meaning; when otherwise, they have none*. And eke-
where3 he says 'admitting (what is not true) that the word right
can have a meaning without reference to utility*. Yet, as
Sidgwick points out,3 'when Bentham explains {Principles of
Morals and Legislation, Chap, I,§ i, note) that his fundamental
principle "states the greatest happiness of all those whose
interest is in question as being the right and proper end of
human action", we cannot understand him really to mean by
the word "right" "conducive to the general happiness"; for
Principles of Morals and Legitlatiant Ch. I,§ 10. * ib.§ '4- "°.
MetkoJ$ of Ethia, cd. 7, 3,6 o.
the proposition that it is conducive to general happiness to
take general happiness as an end of action, though not exactly
a tautology, can hardly serve as the fundamental principle of a
moral system*. Bentham has evidently not made up his mind
clearly whether he thinks that 'right* means 'productive of the
general happiness', or that being productive of the general
happiness is what makes right acts right; and would very
likely have thought the difference unimportant. Mill does not
so far as I know discuss the question whether right is definable.
He states his creed in the form "actions are right in proportion
as they tend to promote happiness',1 where the claim that is
made is not that this is what 'right* means, but that this is the
other characteristic in virtue of which actions that are right are
right. And Sidgwick says2 that the meaning of'right* or 'ought*
'is too elementary to admit of any formal definition*, and ex-
pressly repudiates3 the view that 'right* means 'productive of
any particular sort of result*.
The most deliberate claim that 'right' is definable as 'pro-
ductive of so and so* is made by Prof. G. E. Moore, who claims
in Principle Ethica that 'right* means 'productive of the greatest
possible good'. Now it has often been pointed out against
hedonism, and by no one more clearly than by Professor Moore,
that the claim that 'good* just means 'pleasant* cannot seriously
be maintained; that while it may or may not be true that the
only things that are good are pleasant, the statement that the
good is just the pleasant is a synthetic, not an analytic proposi-
tion; that the words 'good* and 'pleasant* stand for distinct
qualities, even if the things that possess the one are precisely
the things that possess the other. If this were not so, it would
not be intelligible that the proposition 'the good is just the
pleasant* should have been maintained on the one hand, and
denied on the other, with so much fervour; for we do not fight
for or against analytic propositions; we take them for granted.
Must not the same claim be made about the statement "being
right means being an act productive of the greatest good pro-
ducible in the circumstances' ? Is it not plain on reflection that
Utilitarianism, copyright eds,, 9. *• Methods ofEthict, eel. 7, jj.
» ik 25-6.
this is not what we mean by right, even if it be a true statement
about what is right? It seems clear for instance that when an
ordinary man says it is right to fulfil promises he is not in the
least thinking of the total consequences of such an act, about
which he knows and cares little or nothing. 'Ideal utilitaria-
nism* 1 is, it would appear, plausible only when it is understood
not as an analysis or definition of the notion of "right* but as a
statement that all acts that are right, and only these, possess
the further characteristic of being productive of the best pos-
sible consequences, and are right 'because they possess this
other characteristic.
If I am not mistaken, Professor Moore has moved to this
position, from the position that 'right* is anafysaUe into 'pro-
ductive of the greatest possible good*. In Prtacipia Ethka the
latter position is adopted: e.g. 'This use of "right", as denoting
what is good as a means, whether or not it is also good as an
end, is indeed the use to which I shall confine the word*.2 'To
assert that a certain line of conduct is, at a given time, absolutely
right or obligatory, is obviously to assert that more good or
less evil will exist in the world, if it be adopted, than if anything
else be done instead.'3 'To ask what kind of actions one ought
to perform, or what kind of conduct is right, is to ask what kind
of effects such action and conduct will produce,.. What I wish
first to point out is that "right" does and can mean nothing but
"cause of a good result", and is thus always identical with
"useful"... That the assertion "I am morally bound to perform
this action" is identical with the assertion "this action will pro-
duce the greatest possible amount of good in the Universe"
has already been briefly shewn ...; but it is important to insist
that this fundamental point is demonstrably certain... . Our
"duty", therefore, can only be defined as that action, which
will cause more good to exist in the Universe than any possible
alternative. And what is "right" or "morally permissible"
only differs from this, as what will not cause less good than any
possible alternative,' 4.
I use this as a well-known way of referring to Professor Moore's view, 'Aga-
thistie utilitarianism* would indicate more distinctly the difference between it and
hedonistic utilitarianism. p. 18, ' p. ij.
* pp. 146-8. Cf. also pp. 167, 169, 180-1.
In his later book, Ethics^ Professor Moore seems to have
come to adopt the other position, though perhaps not quite
unequivocally. On page 8 he names as one of the 'more funda-
mental questions* of ethics the question 'what, after all, is it
that we mean to say of an action when we say that it is right or
ought to be done? * Here it is still suggested that 'right' is per-
haps analysable or definable. But to this question Ethics no-
where distinctly offers an answer, and on page 9 we find, 'Can
we discover any single reason, applicable to all right actions
equally, which is, in every case, the reason'why an action is
right, when it is right ?' This is the question which Professor
Moore in fact sets himself to answer. But the reason for an
action's being right is evidently not the same thing as its rig&t-
ness, and Professor Moore seems already to have passed to the
view that productivity of maximum good is not the definition
of 'right* but another characteristic which underlies and accounts
for the rightness of right acts. Again, he describes hedonistic
utilitarianism as asking, 'can we discover any characteristic,
over and above the mere fact that they are right, which belongs
to absolutely all voluntary actions which are right, and which
at the same time does not belong to any except those which are
right?'I This is the question which he describes hedonism as
essentially answering, and since his own view differs from
hedonism not in logical form but just by the substitution of
'good* for "pleasure", his theory also seems to be essentially an
answer to this question, i.e. not to the question what is right-
ness but to the question what is the universal accompaniment
and, as he is careful to add,2 the necessitating ground of right-
ness. Again, he describes hedonistic utilitarianism as giving us
'a criterion, or test, or standard by which we could discern
with regard to any action whether it is right or wrong*.3 And
similarly, I suppose, he regards his own theory as offering a
different criterion of rightness. But obviously a criterion of
rightness is not rightness itself. And, most plainly of all, he
says, 'It is indeed quite plain, I think, that the meaning of the two
words" ('duty* and 'expediency', the latter being equivalent to
'tendency to produce the maximum good*) 'is not the samej
p. 17. * PP- 44, 54- * 9- 43-
for, if it were, then it would be a mere tautology to say that it is
always our duty to do what will have the best possible conse-
quences'.1 If we contrast this with Principia Ethka, page 169,
'if I ask whether an action is really my duty or really expedient,
the predicate of which I question the applicability to the action
in question is precisely the same*, we see how much Professor
Moore has changed his position, and changed it in the direction
in which, as I have been urging, it must be changed if it is to be
made plausible. And if it is clear that 'right' does not mean
'productive of the greatest possible good', it is a fortiori clear
that it does not mean 'productive of the greatest possible plea-
sure, for the agent or for mankind*, but that productivity of
the greatest possible pleasure for the agent or for mankind is at
most the ground of the lightness of acts, Tightness itself being
admitted to be a distinct characteristic, and one which utili-
tarianism does not claim to define.
But there are theories other than utilitarianism which claim
to define 'right*. It would be tedious to try to refute all such
theories. With regard to many of them 2 it seems to be enough
to ask one's readers whether it is not clear to them on reflection
that the proposed definition of 'right* bears in fact no resem-
blance to what they mean by 'right'. But there is one group of
theories to which some reference should be made, viz. those
that give what may be called a subjective theory of 'right',
that identify the Tightness of an act with its tendency to produce
either some feeling or some opinion in the mind of some one
who contemplates it. This type of theory has been dealt with
very thoroughly by Professor Moore,3 and I should have little
or nothing to add to his convincing refutation. But such theories
are perhaps even more prevalent with regard to 'good* than
to 'right', and in my fourth chapter I discuss them at some
length. I would ask my readers to read the argument there
offered, and to reflect whether the refutation I offer 4 of
subjective accounts of 'good' does not apply with equal force
to subjective accounts of'right*.
P. 173.
* e.g. the evolutionary theory which identifies 'right' with 'conducive to life*.
' Ettict, Chs, j, 4, * pp. 80-104.
Any one who is satisfied that neither the subjective theories
of the meaning of 'right', nor what is far the most attractive of
the attempts to reduce it to simpler objective elements, is cor-
rect, will probably be prepared to agree that 'right* is an irre-
ducible notion.
Nor is this result impugned by inquiries into the historical
development of our present moral notions from an earlier state
of things in which 'what is right' was hardly disentangled from
'what the tribe ordains*. The point is that we can now see
clearly that 'right* does not mean 'ordained by any given
society*. And it may be doubted whether even primitive men
thought that it did. Their thoughts about what in particular
was right were to a large extent limited by the customs and
sanctions of their race and age. But this is not the same as
to say that they thought that 'right* just meant 'what my race
and age ordains'. Moral progress has been possible just because
there have been men in all ages who have seen the difference
and have practised, or at least preached, a morality in some
respects higher than that of their race and age. And even the
supporters of the lower morality held, we may suspect, that
their laws and customs were in accordance with a 'right* other
than themselves, 'It is the custom' has been accompanied by
'the custom is right*, or 'the custom is ordained by some one
who has the right to command'. And if human consciousness
is continuous, by descent, with a lower consciousness which
had no notion of right at all, that need not make us doubt that
the notion is an ultimate and irreducible one, or that the right-
ness (prima facie)1 of certain types of act is self-evident; for the
nature of the self-evident is not to be evident to every mind
however undeveloped, but to be apprehended directly by
minds which have reached a certain degree of maturity, and
for minds to reach the necessary degree of maturity the de-
velopment that takes place from generation to generation is as
much needed as that which takes place from infancy to adult life.
In this connexion it may be well to refer briefly to a theory
which has enjoyed much popularity, particularly in France—
the theory of the sociological school of Durkheim and Levy-
' For this qualification cf. pp. 19-20.
Bruhl, which seeks "to replace moral philosophy by the 'science
des mceurs', the historical and comparative study of the moral
beliefs and practices of mankind. It would be foolish to deny
the value of such a study, or the interest of many of the facts it
has brought to light with regard to the historical origin of
many such beliefs and practices. It has shown with success that
many of the most strongly felt repulsions towards certain types
of conduct are relics of a bygone system of totems and fetishes,
their connexion with which is little suspected by those who
feel them. What must be denied is the capacity of any such
inquiry to take the place of moral philosophy. The attitude of
the sociological school towards the systems of moral belief
that they find current in various ages and races is a curiously
inconsistent one. On the one hand we are urged to accept an
existing code as something analogous to an existing law of
nature, something not to be questioned or criticized but to be
accepted and conformed to as part of the given scheme of
things j and on this side the school is able sincerely to proclaim
itself conservative of moral values, and is indeed conservative
to the point of advocating the acceptance in full of conven-
tional morality. On the other hand, by showing that any given
code is the product partly of bygone superstitions and partly
of out-of-date utilities, it is bound to create in the mind of
any one who accepts its teaching (as it presupposes in the mind
of the teacher) a sceptical attitude towards any and every given
code. In fact the analogy which it draws between a moral code
and a natural system like the human body (a favourite com-
parison) is an entirely fallacious one. By analysing the con-
stituents of the human body you do nothing to diminish the
reality of the human body as a given fact, and you learn much
which will enable you to deal effectively with its diseases. But
beliefs have the characteristics which bodies have not, of being
true or false, of resting on knowledge or of being the product
of wishes, hopes, and fears; and in so far as you can exhibit
them as being the product of purely psychological and non-
lojpcal causes of this sort, while you leave intact the fact that
many people hold such opinions you remove their authority
and their claim to be carried out in practice.
It is often said, in criticism of views such as those of the
sociological school, that the question of the validity of a moral
code is quite independent of the question of its origin. This
does not seem to me to be true. An inquiry into the origin of a
judgement may have the effect of establishing its validity.
Take, for instance, the judgement that the angles of a triangle
are equal to two right angles. We find that the historical origin
of this judgement lies in certain pre-existing judgements which
are its,premisses, plus the exercise of a certain activity of in-
ferring. Now if we find that these pre-existing judgements
were really instances of knowing, and that the inferring was
also really knowing—was the apprehension of a necessary
connexion—our inquiry into the origin of the judgement in
question will have established its validity. On the other hand,
if any one can show that A holds actions of type S to be wrong
simply because (for instance) he knows such actions to be
forbidden by the society he lives in, he shows that A has no
real reason for believing that such actions have the specific
quality of wrongness, since between being forbidden By the
community and being wrong there is no necessary connexion.
He does not, indeed, show the belief to be untrue, but he shows
that A has no sufficient reason for holding it trae^ and in this
sense he undermines its validity.
This is, in principle, what the sociological school attempts
to do. According to this school, or rather according to its
principles if consistently carried out, no one moral code is any
truer, any nearer to the apprehension of an objective moral
truth, than any other; each is simply the code that is necessitated
by the conditions of its time and place, and is that which most
completely conduces to the preservation of the society that
accepts it. But the human mind will not rest content with such
a view. It is not in the least bound to say that there has been
constant progress in morality, or in moral belief. But it is
competent to see that the moral code of one race or age is in
certain respects inferior to that of another. It has in feet an
a priori insight into certain broad principles of morality, and it
can distinguish between a more and a less adequate recognition
of these principles. There are not merely so many moral codes
which can be described and whose vagaries can be traced to
historical causes; there is a system of moral truth, as objective
as all truth must be, which, and whose implications, we are
interested in discovering; and from the point of view of this,
the genuinely ethical problem, the sociological inquiry is simply
beside the mark. It does not touch the questions to which we
most desire answers.1
For a lucid atrf up to a point appreciative account of the sociological school, and
a penetrating criticism of its deficiencies, see ch, 4 of M. D, Parodfs Lt PraUimt
Moreltt la Paufe Gmteatforaine,
HE real point at issue between hedonism and utilitafianism
T on the one hand and their opponents on the other is not
whether 'right' means 'productive of so and so'; for it cannot
with any plausibility be maintained that it does. The point at
issue is that to which we now pass, viz. whether there is any
general character which makes right acts right, and if so, what
it is. Among the main historical attempts to state a single
characteristic of all right actions which is the foundation of their
lightness are those made by egoism and utilitarianism. But I
do not propose to discuss these, not because the subject is un-
important, but because it has been dealt with so often and
so well already, and because there has come to be so much
agreement among moral philosophers that neither of these
theories is satisfactory. A much more attractive theory has been
put forward by Professor Moore: that what makes actions
right is that they are productive of more good than could have
been produced by any other action open to the agent.1
This theory is in fact the culmination of all the attempts to
base tightness on productivity of some sort of result. The
first form this attempt takes is the attempt to base tightness on
conduciveness to the advantage or pleasure of the agent. This
theory comes to grief over the fact, which stares us in the face,
that a great part of duty consists in an observance of the rights
and a furtherance of the interests of others, whatever the cost to
ourselves may be. Plato and others may be right in holding
that a regard for the rights of others never in the long run
involves a loss of happiness for the agent, that 'the just life
profits a man*. But this, even if true, is irrelevant to the right-
ness of the act. As soon as a man does an action because he
thinks he will promote his own interests thereby, he is acting
not from a sense of its Tightness but from self-interest.
I take the theory which, as I have tried to show, seems to be put forward in Ethics
rather than the earlier and less plausible theory put forward in Pruttipis Ethkn. For
the difference, cf, my pp. 8-n.
To the egoistic theory hedonistic utilitarianism supplies a
much-needed amendment. It points out correctly that the fact
that a certain pleasure will be enjoyed by the agent is no reason
why he ought to bring it into being rather than an equal or
greater pleasure to be enjoyed by another, though, human
nature being what it is, it makes it not unlikely that he will try
to bring it into being. But hedonistic utilitarianism in its turn
needs a correction. On reflection it seems clear that pleasure
is not the only thing in life that we think good in itself, that
for instance we think the possession of a good character, or an
intelligent understanding of the world, as good or better. A
great advance is made by the substitution of "productive of
the greatest good* for 'productive of the greatest pleasure'.
Not only is this theory more attractive than hedonistic
utilitarianism, but its logical relation to that theory is such that
the latter could not be true unless it were true, while it might
be true though hedonistic utilitarianism were not. It is in fact
one of the logical bases of hedonistic utilitarianism. For the
view that what produces the maximum pleasure is right has for
its bases the views (i) that what produces the maximum good
is right, and (2) that pleasure is the only thing good in itself.
If they were not assuming that what produces the maximum
good is right, the utilitarians* attempt to show that pleasure is
the only thing good in itself, which is in fact the point they
take most pains to establish, would have been quite irrelevant
to their attempt to prove that only what produces the maximum
pleasure is right. If, therefore, it can be shown that productivity
of the maximum good is not what makes all right actions right,
we shall a fortiori have refuted hedonistic utilitarianism.
When a plain man fulfils a promise because he thinks he
ought to do so, it seems clear that he does so with no thought
of its total consequences, still less with any opinion that these
are likely to be the best possible. He thinks in fact much more
of the past than of the future. What makes him think it right
to act in a certain way is the fact that he has promised to do so
—that and, usually, nothing more. That his act will produce the
best possible consequences is not his reason for calling it right.
What lends colour to the theory we are examining, then, is not
the actions (which form probably a great majority of our
actions) in which some such reflection as 'I have promised' is
the only reason we give ourselves for thinking a certain action
right, but the exceptional cases in which the consequences of
fulfilling a promise (for instance) would be so disastrous to
others that we judge it right not to do so. It must of course be
admitted that such cases exist. If I have promised to meet a
friend at a particular time for some trivial purpose, I should
certainly think myself justified in breaking my engagement if
by doing so I could prevent a serious accident or bring, relief
to the victims of one. And the supporters of the view we are
examining hold that my thinking so is due to my thinking that
I shall bring more good into existence by the one action than
by the other, A different account may, however, be given of
the matter, an account which will, I believe, show itself to be
the true one. It may be said that besides the duty of fulfilling
promises I have and recognize a duty of relieving distress,1 and
that when I think it right to do the latter at the cost of not doing
the former, it is not because I think I shall produce more good
thereby but because I think it the duty which is in the circum-
stances more of a duty. This account surely corresponds much
more closely with what we really think in such a situation. If,
so far as I can see, I could bring equal amounts of good into
being by fulfilling my promise and by helping some one to
whom I had made no promise, I should not hesitate to regard
the former as my duty. Yet on the view that what is right is
right because it is productive of the most good I should not so
regard 5t.
There are two theories, each in its way simple, that offer a
solution of such cases of conscience. One is the view of Kant,
that there are certain duties of perfect obligation, such as those
of fulfilling promises, of paying debts, of telling the truth,
which admit of no exception whatever in favour of duties of
imperfect obligation, such as that of relieving distress. The
other is the view of, for instance, Professor Moore and Dr.
Rashdall, that there is only the duty of producing good, and
These are not strictly speaking duties, but things that tend to be our duly, or
prima-facie duties. Cf, pp. 19-10.
that all 'conflicts of duties* shoukTbe resolved by asking *by
which action will most good be produced?' But it is more
important that our theory fit the facts than that it be simple,
and the account we have given above corresponds (it seems to
me) better than either of the simpler theories with what we
really think, viz, that normally promise-keeping, for example,
should come before benevolence, but that when and only when
the good to be produced by the benevolent act is very great
and the promise comparatively trivial, the act of benevolence
becomes our duty.
In fact the theory of'ideal utilitarianism*, if I may for brevity
refer so to the theory of Professor Moore, seems to simplify
unduly our relations to our fellows. It says, in effect, that the
only morally significant relation in which my neighbours stand
to me is that of being possible beneficiaries by my action,1
They do stand in this relation to me, and this relation is morally
significant. But they may also stand to me in the relation of
promisee to promiser, of creditor to debtor, of wife to husband,
of child to parent, of friend to friend, of fellow countryman
to fellow countryman, and the like; and each of these relations
is the foundation of a prima facie duty, which is more or less
incumbent on rne according to the circumstances of the case.
When I am in a situation, as perhaps I always am, in which
more than one of these prima facie duties is incumbent on me,
what I have to do is to study the situation as fully as I can until
I form the considered opinion (it is never more) that in the
circumstances one of them is more incumbent than any other;
then I am bound to think that to do this pritna facie duty is my
duty sans phrase in the situation,
I suggest *prima facie duty* or 'conditional duty* as a brief
way of referring to the characteristic (quite distinct from that
of being a duty proper) which an act has, in virtue of being of
a certain kind (e.g. the keeping of a promise), of being an act
which would be a duty proper if it were not at the same time
of another kind which is morally significant. Whether an act
Some will think it, apart from other considerations, a sufficient refutation of this
view to point out that I also stand in that relation to myself, so that for this view the
distinction of oneself from others is morally insignificant.
is a duty proper or actual duty depends on all the morally
significant kinds it is an instance of. The phrase * prima facie
duty* must be apologized for, since (i) it suggests that what we
are speaking of is a certain kind of duty, whereas it is in fact
not a duty, but something related in a special way to duty.
Strictly speaking, we want not a phrase in which duty is quali-
fied by an adjective, but a separate noun, (2) 'Prima'facie sug-
gests that one is speaking only of an appearance which a moral
situation presents at first sight, and which may turn out to be
illusory; whereas what I am speaking of is an objective fact
involved in the nature of the situation, or more strictly in an
element of its nature, though not, as duty proper does, arising
from its whole nature. * I can, howeveiythink of no term which
fully meets the case. 'Claim* has been suggested by Professor
Prichard. The word 'claim* has the advantage of being quite
& familiar one in this connexion, and it seems to cover much of
the ground. It wouldbe quite natural to say, *a person to whom
I have made a promise has a claim on me', and also, *a person
whose distress I could relieve (at the cost of breaking the
promise) has a claim on me". But (i) while 'claim1 is appro-
priate from their point of view, we want a word to express the
corresponding fact from the agent's point of view—the fact
of his being subject to claims that can be made against him;
and ordinary language provides us with no such correlative to
'claim*. And (2) (what is more important) 'claim' seems in-
evitably to suggest two persons, one of whom might make a
claim on the other; and while this covers the ground of social
duty, it is inappropriate in the case of that important part of
duty which is the duty of cultivating a certain kind of character
in oneself. It would be artificial, I think, and at any rate meta-
phorical, to say that one's character has a claim on oneself. *
There is nothing arbitrary about these prima facie duties.
Each rests on a definite circumstance which cannot seriously be
held to be without moral significance. Of prima facie duties
I suggest, without claiming completeness or finality for it, the
following division,1
* I should make it plain at this stage that I am assuming the correctness of some of
our main convictions as to prima facie duties, or, more strictly, am claiming that we
(i) Some duties rest on previous acts of my own. These
duties seem to include two kinds, (a) those resting on a promise
or what may fairly be called an implicit promise, such as the
implicit undertaking not to tell lies which seems to be implied
in the act of entering into conversation (at any rate by civilized
men), or of writing books that purport to be history and not
fiction. These may be called the duties of fidelity, (A) Those
resting on a previous wrongful act. These may be called the
duties of reparation. (2) Some rest on previous acts of other
men, ie, services done by them to me. These may be loosely
described as the duties of gratitude.1 (3) Some rest on the fact
or possibility of a distribution of pleasure or happiness (or of
the means thereto) which is not in accordance with the merit
of the persons concerned; in such cases there arises a duty to
upset or prevent such a distribution. These are the duties of
justice. (4) Some rest on the mere fact that there are other
beings in the world whose condition we can make better in
respect of virtue, or of intelligence, or of pleasure. These are
the duties of beneficence, (y) Some rest on the fact that we can
improve our own condition in respect of virtue or of intelli-
gence. These are the duties of self-improvement. (6) I think
that we should distinguish from (4) the duties that may be
summed up under the title of 'not injuring others'. No doubt
to injure others is incidentally to fail to do them good; but it
seems to me clear that non-maleficence is apprehended as a
duty distinct from that of beneficence, and as a duty of a more
stringent character. It will be noticed that this alone among
the types of duty has been stated in a negative way." An attempt
might no doubt be made to state this duty, like the others, in
a positive way. It might be said that it is really the duty to
how them to be true. To me it seems as self-evident as anything could be, that to
make a promise, for instance, is to create a moral claim on us in someone else. Many
readers will perhaps say that they do not know this to be true. If so, I certainly caoitot
prove it to them; I can only ask them to reflect again, in die hope that they will ulti-
mately agree that they also know it to be true. The main moral convictions of the
plain man seem to me to be, not opinions which it is for philosophy to prove or dis-
prove, but knowledge from the start; and in my own case I seem to find little difficulty
in distinguishing these essential convictions Iron* other moral convictions which 1
also have, which are merely fallible opinions based ort an imperfect study of the work*
ing for good or evil of certain institutions or types of action.
' For a needed correction of this statement, cf. pp. M-J.
prevent ourselves from acting either from an inclination to
harm others or from an inclination to seek our own pleasure, in
doing which we should incidentally harm them. But on reflection
it seems clear that the primary duty here is the duty not to
harm others, this being a duty whether or not we have an
inclination that if followed would lead to our harming them;
and that when we have such an inclination the primary duty
not to harm others gives rise to a consequential duty to resist
the inclination. The recognition of this duty of non-maleficence
is the first step on the way to the recognition of the duty of
beneficence; and that accounts for the prominence of the com-
mands 'thou shalt not kill', 'thou shah not commit adultery',
'thou shalt not steal*, 'thou shalt not bear false witness', in so
early a code as the Decalogue. But even when we have come
to recognize the duty of beneficence, it appears to me that the
duty of non-maleficence is recognized as a distinct one, and as
prima facie more binding. We should not in general consider
it justifiable to kill one person in order to keep another alive,
or to steal from one in order to give alms to another.
The essential defect of the 'ideal utilitarian* theory is that
it ignores, or at least does not do full justice to, the highly
personal character of duty. If the only duty is to produce the
maximum of good, the question who is to have the good—•
whether it is myself, or my benefactor, or a person to whom I
have made a promise to confer that good on him, or a mere
fellow man to whom I stand in no such special relation—
should make no difference to my having a duty to produce that
good. But we are all in fact sure that it makes a vast difference.
One or two other comments must be made on this pro-
visional list of the divisions of duty, (i) The nomenclature is
not strictly correct. For by 'fidelity* or 'gratitude' we mean,
strictly, certain states of motivation; and, as I have urged, it is
not our duty to have certain motives, but to do certain acts.
By 'fidelity', for instance, is meant, strictly, the disposition to
fulfil promises and implicit promises because we have made them,
We have no general word to cover the actual fulfilment of
promises and implicit promises irrespective of motive; and I use
'fidelity', loosely but perhaps conveniently, to fill this gap. So
too I use 'gratitude* for the returnitig of services, irrespective
of motive. The term 'justice* is not so much confined, in ordi-
nary usage, to a certain state of motivation, for we should
often talk of a man as acting justly even when we did not think
his motive was the wish to do what was just simply for the sake
of doing so. Less apology is therefore needed for our use of
'justice' in this sense. And I have used the word "beneficence*
rather than 'benevolence', in order to emphasize the fact that
it is our duty to do certain things, and not to do them from
certain motives.
(2) If the objection be made, that this catalogue of the main
types of duty is an unsystematic one resting on no logical
principle, it may be replied, first, that it makes no claim to being
ultimate. It is a prima facie classification of the duties which
reflection on our moral convictions seems actually to reveal.
And if these convictions are, as I would claim that they are, of
the nature of knowledge, and if I have not misstated them, the
list will be a list of authentic conditional duties, correct as far
as it goes though not necessarily complete. The list of goods
put forward by the rival theory is reached by exactly the same
method—the only sound one in the circumstances—viz, that
of direct reflection on what we really think. Loyalty to the
facts is worth more than a symmetrical architectonic or a hastily
reached simplicity. If further reflection discovers a perfect
logical basis for this or for a better classification, so much the
(3) It may, again, be objected that our theory that there are
these various and often conflicting types of prima facie duty
leaves us with no principle upon which to discern what is our
actual duty in particular circumstances. But this objection is
not one which the rival theory is in a position to bring forward.
For when we have to choose between the production of two
heterogeneous goods, say knowledge and pleasure, the 'ideal
utilitarian* theory can only fall back on an opinion, for which
no logical basis can be offered, that one of the goods is the
greater; and this is no better than a similar opinion that one of
two duties is the more urgent. And again, when we consider
the infinite variety of the effects of our actions in the way of
pleasure, it must surely be admitted that the claim which
hedonism sometimes makes, that it offers a readily applicable cri-
terion of right conduct, is quite illusory.
I am unwilling, however, to content myself with an argu-
mentum ad Aominem, and I would contend that in principle
there is no reason to anticipate that every act that is our duty is
so for one and the same reason. Why should two sets of cir-
cumstances, or one set of circumstances, not possess different
characteristics, any one of which makes a certain act omprima
fade duty? When I ask what it is that makes me in certain
cases sure that I have a prima facie duty to do so and so, I find
that it lies in the fact that I have made a promise; when I ask
the same question in another case, I find the answer lies in the
fact that I have done a wrong. And if on reflection I find (as I
think I do) that neither of these reasons is reducible to the other,
I must not on any a priori ground assume that such a reduction
is possible.
An attempt may be made to arrange in a more systematic
way the main types of duty which we have indicated. In the
first place it seems self-evident that if there are things that are
intrinsically good, it is prima facie a duty to bring them into
existence rather than not to do so, and to bring as much of them
into existence as possible. It will be argued in our fifth chapter
that there are three main things that are intrinsically good—
virtue, knowledge, and, with certain limitations, pleasure. And
since a given virtuous disposition, for instance, is equally good
whether it is realized in myself or in another, it seems to be my
duty to bring it into existence whether in myself or in another.
So too with a given piece of knowledge.
The case of pleasure is difficult; for while we clearly recog-
nize a duty to produce pleasure for others, it is by no means so
clear that we recognize a duty to produce pleasure for our-
selves. This appears to arise from the following facts. The
thought of an act as our duty is one that presupposes a certain
amount of reflection about the act; and for that reason does not
normally arise in connexion with acts towards which we are
already impelled by another strong impulse. So far, the cause
of our not thinking of the promotion of our own pleasure as a
duty is analogous to the cause which usually prevents a highly
sympathetic person from thinking of the promotion of the
pleasure of others as a duty. He is impelled so strongly by
direct interest in the well-being of others towards promoting
their pleasure that he does not stop to ask whether it is his duty
to promote it; and we are all impelled so strongly towards the
promotion of our own pleasure that we do not stop to ask
whether it is a duty or not. But there is a further reason why
even when we stop to think about the matter it does not usually
present itself as a duty: viz. that, since the performance of most
of our duties involves the giving up of some pleasure that we
desire, the doing of duty and the getting of pleasure for our-
selves come by a natural association of ideas to be thought of as
incompatible things. This association of ideas is in the main
salutary in its operation, since it puts a check on what but for
it would be much too strong, the tendency to pursue one's
own pleasure without thought of other considerations. Yet
if pleasure is good, it seems in the long run clear that it is right
to get it for ourselves as well as to produce it for others, when
this does not involve the failure to discharge some more strin-
gent prims facie duty. The question is a very difficult one, but
it seems that this conclusion can be denied only on one or other
of three grounds: (i) that pleasure is not prima facie good (i.e.
good when it is neither the actualization of a bad disposition
nor undeserved), (2) that there is no primafacie duty to produce
as much that is good as we can, or (3) that though there is a
prima facie duty to produce other things that are good, there
is no prima facie duty to produce pleasure which will be en-
joyed by ourselves. I give reasons later" for not accepting the
first contention. The second hardly admits of argument but
seems to me plainly false. The third seems plausible only if we
hold that an act that is pleasant or brings pleasure to ourselves
must for that reason not be a duty; and this would lead to para-
doxical consequences, such as that if a man enjoys giving
pleasure to others or working for their moral improvement,
it cannot be his duty to do so. Yet it seems to be a very stub-
bom fact, that in our ordinary consciousness we are not aware
" PP. »3J-*-
of a duty to get pleasure for ourselves; and by way of partial
explanation of this I may add that though, as I think, one's
own pleasure is a good and there is a duty to produce it, it is
only if we think of our own pleasure not as simply our own
pleasure, but as an objective good, something that an impartial
spectator would approve, that we can think of the getting it as
a duty; and we do not habitually think of it in this way, *
If these contentions are right, what we have called the duty
of Beneficence and the duty of self-improvement rest on the
same ground. No different principles of duty are involved in
the two cases. If we feel a special responsibility for improving
our own character rather than that of others, it is not because
a special principle is involved, but because we are aware that
the one is more under our control than the other. It was on
this ground that Kant expressed the practical law of duty in
the form 'seek to make yourself good and other people happy*.
He was so persuaded of the internality of virtue that he re-
garded any attempt by one person to produce virtue in another
as bound to produce, at most, only a counterfeit of virtue, the
doing of externally right acts not from the true principle of
virtuous action but out of regard to another person. It must
be admitted that one roan cannot compel another to be virtuous;
compulsory virtue would just not be virtue. But experience
clearly shows that Kant overshoots the mark when he contends
that one man cannot do anything to promote virtue in another,
to bring such influences to bear upon him that his own response
to them is more likely to be virtuous than his response to other
influences would have been. And our duty to do this is not
different in kind from our duty to improve our own characters.
It is equally clear, and clear at an earlier stage of moral de-
velopment, that if there are things that are bad in themselves
we ought, prima facie, not to bring them upon others; and on
this fact rests the duty of non-maleficence.
The duty of justice is particularly complicated, and the word
is used to cover things which are really very different—things
such as the payment of debts, the reparation of injuries done by
oneself to another, and the bringing about of a distribution of
happiness between other people in proportion to merit. I use
the word to denote only the last of these three. In die fifth
chapter I shall try to show that besides the three (compara-
tively) simple goods, virtue, knowledge, and pleasure, there is
a more complex good, not reducible to these, consisting in the
proportionment of happiness to virtue. The bringing of this
about is a duty which we owe to all men alike, though it may
be reinforced by special responsibilities that we have under-
taken to particular men. This, therefore, with beneficence and
self-improvement, comes under the general principle that we
should produce as much good as possible, though the good
here involved is different in kind from any other.
But besides this general obligation, there are special obliga-
tions. These may arise, in the first place, incidentally, from
acts which were not essentially meant to create such an obliga-
tion, but which nevertheless create it. From the nature of the
case such acts may be of two kinds—the infliction of injuries
on others, and the acceptance of benefits from them. It seems
clear that these put us under a special obligation to other men,
and that only these acts can do so incidentally. From these
arise the twin duties of reparation and gratitude,
And finally there are special obligations arising from acts
the very intention of which, when they were done, was to put
us under such an obligation. The name for such acts is 'pro-
mises'; the name is wide enough if we are willing to include
under it implicit promises, i.e. modes of behaviour in which
without explicit verbal promise we intentionally create an ex-
pectation that we can be counted on to behave in a certain
way in the interest of another person.
These seem to be, in principle, all the ways in which prima
facie duties arise. In actual experience they are compounded
together in highly complex ways. Thus, for example, the duty
of obeying the laws of one's country arises partly (as Socrates
contends in the Crito) from the duty of gratitude for the bene-
fits one has received from it; partly from the implicit promise
to obey which seems to be involved in permanent residence in a
country whose laws we know we are expected to obey, and still
more clearly involved when we ourselves invoke the protection
of its laws (this is the truth underlying the doctrine of the
social contract); and partly (if we are fortunate in our country)
from the fact that its laws are potent instruments for the general
Or again, the sense of a general obligation to bring about
(so far as we can) a just apportionment of happiness to merit is
often greatly reinforced by the fact that many of the existing
injustices are due to a social and economic system which we
have, not indeed created, but taken part in and assented to; the
duty of justice is then reinforced by the duty of reparation.
It is necessary to say something by way of clearing up the
relation between prima facie duties and the actual or absolute
duty to do one particular act in particular circumstances. If,
as almost all moralists except Kant are agreed, and as most
plain men think, it is sometimes right to tell a lie or to break a
promise, it must be maintained that there is a difference between
prima facie duty and actual or absolute duty. When we think
ourselves justified in breaking, and indeed morally obliged to
break, a promise in order to relieve some one's distress, we do
not for a moment cease to recognize a prima facie duty to keep
our promise, and this leads us to feel, not indeed shame or
repentance, but certainly compunction, for behaving as we do;
we recognize, further, that it is our duty to make up somehow
to the promisee for the breaking of the promise. We have to
distinguish from the characteristic of being our duty that of
tending to be our duty. Any act that we do contains various
elements in virtue of which it falls under various categories.
In virtue of being the breaking of a promise, for instance, it
tends to be wrong; in virtue of being an instance of relieving
distress it tends to be right. Tendency to be one's duty may be
called a parti-resultant attribute, i.e. one which belongs to an
act in virtue of some one component in its nature. Being one's
duty is a tori-resultant attribute, one which belongs to an act
in virtue of its whole nature and of nothing less than this.l This
distinction between parti-resultant and toti-resukant attributes
is one which we shall meet in another context also.2
Another instance of the same distinction may be found in
the operation of natural laws. Qua subject to the force of
But ef. the qualification in p. jj, a. a. * Cf. pp. 1*1-3.
gravitation towards some other body, each body tends to move
in a particular direction with a particular velocity; but its actual
movement depends on all the forces to which it is subject. It
is only by recognizing this distinction that we can preserve the
absoluteness of laws of nature, and only by recognizing a
corresponding distinction that we can preserve the absoluteness
of the general principles of morality. But an important difference
between the two cases must be pointed out. When we say that
in virtue of gravitation a body tends to move in a certain way,
we are referring to a causal influence actually exercised on it by
another body or other bodies. When we say that in virtue of
being deliberately untrue a certain remark tends to be wrong,
we are referring to no causal relation, to no relation that involves
succession in time, but to such a relation as connects the various
attributes of a mathematical figure. And if the word 'tendency*
is thought to suggest too much a causal relation, it is better to
talk of certain types of act as being prima facie right or wrong
(or of different persons as having different and possibly con-
flicting claims upon us), than of their tending to be right or
Something should be said of the relation between our appre-
hension of the prima fade Tightness of certain types of act and
our mental attitude towards particular acts, It is proper to use
the word 'apprehension* in the former case and not in the latter.
That an act, qua fulfilling a promise, or qua effecting a just
distribution of good, or qua returning services rendered, or
qua promoting the good of others, or qua promoting the virtue
or insight of the agent, is prima facie right, is self-evident; not
in the sense that it is evident from the beginning of our lives,
or as soon as we attend to the proposition for the first time, but
in the sense that when we have reached sufficient mental maturity
and have given sufficient attention to the proposition it is evi-
dent without any need of proof, or of evidence beyond itself,
It is self-evident just as a mathematical axiom, or the validity of
a form of inference, is evident. The moral order expressed in
these propositions is just as much part of the fundamental
nature of the universe (and, we may add, of any possible uni-
verse in which there were moral agents at all) as is the spatial
or numerical structure expressed in the axioms of geometry or
arithmetic. In OUT confidence that these propositions are true
there is involved the same trust in our reason that is involved
in our confidence in mathematics; and we should have no
justification for trusting it in the latter sphere and distrusting
it in the former. In both cases we are dealing with propositions
that cannot be proved, but that just as certainly need no
proof, *
Some of these general principles of prima facie duty may
appear to be open to criticism. It may be thought, for example,
that the principle of returning good for good is a falling off
from the Christian principle, generally and rightly recognized
as expressing the highest morality, of returning good for evil.
To this it may be replied that I do not suggest that there is a
principle commanding us to return good for good and for-
bidding us to return good for evil, and that I do suggest ^hat
there is a positive duty to seek the good of all men. What I
maintain is that an act in which good is returned for good is
recognized as specially binding on us just because it is of that
character, and that ceteris pariius any one would think it his
duty to help his benefactors rather than his enemies, if he could
not do both; just as it is generally recognized that ceterisparibus
we should pay our debts rather than give our money in charity,
when we cannot do both. A benefactor is not only a man,
calling for our effort on his behalf on that ground, but also our
benefactor, calling for our special effort on that ground.
Our judgements about our actual duty in concrete situations
have none of the certainty that attaches to our recognition of
the general principles of duty. A statement is certain, i.e. is an
expression of knowledge, only in one or other of two cases:
when it is either self-evident, or a valid conclusion from self-
evident premisses. And our judgements about our particular
duties have neither of these characters, (i) They are not self-
evident. Where a possible act is seen to have two characteristics,
in virtue of one of which it is prima facie right, and in virtue of
the other prima facie wrong, we are (I think) well aware that
we are not certain whether we ought or ought not to do it;
that whether we do it or not, we are taking a moral risk. We
come in the long run, after consideration, to think one duty
more pressing than the other, but we do not feel certain that
it is so. And though we do not always recognize that a possible
act has two such characteristics, and though there may be cases
in which it has not, we are never certain that any particular
possible act has not, and therefore never certain that it is right,
nor certain that it is wrong. For, to go no further in the analysis,
it is enough to point out that any particular act will in all
probability in the course of time contribute to the bringing
about of good or of evil for many human beings, and thus have
aprima facie Tightness or wrongness of which we know nothing,
(z) Again, our judgements about our particular duties are not
logical conclusions from self-evident premisses. The only
possible premisses would be the general principles stating their
prima facie Tightness or wrongness qua, having the different
characteristics they do have; and even if we could (as we cannot)
apprehend the extent to which an act will tend on the one hand,
for example, to bring about advantages for our benefactors,
and on the other hand to bring about disadvantages for fellow
men who are not our benefactors, there is no principle by which
we can draw the conclusion that it is on the whole right or on
the whole wrong. In this respect the judgement as to the right-
ness of a particular act is just like the judgement as to the beauty
of a particular natural object or work of art. A poem is, for
instance, in respect of certain qualities beautiful and in respect
of certain others not beautiful; and our judgement as to the
degree of beauty it possesses on the whole is never reached by
logical reasoning from the apprehension of its particular beauties
or particular defects. Both in this and in the moral case we
have more or less probable opinions which are not logically
justified conclusions from the general principles that are recog-
nized as self-evident.
There is therefore much truth in the description of the right
act as a fortunate act. If we cannot be certain that it is right, it
is our good fortune if the act we do is the right act. This con-
sideration does not, however, make the doing of our duty a
mere matter of chance. There is a parallel here between the
doing of duty and the doing of what will be to our personal
advantage. We never know what act will in the long run be to
our advantage. Yet It is certain that we are more likely in
general to secure our advantage if we estimate to the best of
our ability the probable tendencies of our actions in this respect,
than if we act on caprice. And similarly we are more likely to
do our duty if we reflect to the best of our ability on dieprima
facie Tightness or wrongness of various possible acts in virtue
of the characteristics we perceive them to have, than if we act
without reflection. With this greater likelihood we must be
Many people would be inclined to say that the right act for
me is not that whose general nature I have been describing,
viz, that which if I were omniscient I should see to be my duty,
but that which on all the evidence available to me I should
think to be my duty. But suppose that from the state of partial
knowledge in which I think act A to be my duty, I could pass
to a state of perfect knowledge in which I saw act S to be my
duty, should I not say 'act B was the right act for me to do* ?
I should no doubt add 'though I am not to be blamed for doing
act A\ But in adding this, am I not passing from thequestion
'what is right' to the question 'what is morally good*? At the
same time I am not making thefull passage from the one notion
to the other; for in order that the act should be morally good,
or an act I am not to be blamed for doing, it must not merely be
the act which it is reasonable for me to think my duty; it must
also be done for that reason, or from some other morally good
motive. Thus the conception of the right act as the act which
it is reasonable for me to think my duty is an unsatisfactory
compromise between the true notion of the right act and the
notion of the morally good action.*
The general principles of duty are obviously not self-evident
from the beginning of our lives. How do they come to be so?
The answer is, that they come to be self-evident to us just as
mathematical axioms do. We find by experience that this
couple of matches and that couple make four matches,
that this couple of balls on a wire and that couple make
four balls; and by reflection on these and similar discoveries
we come to see that it is of the nature of two and two to
make four. In a precisely similaf way, we see the prima
facie Tightness of an act which would be the fulfilment of
a particular promise, and of another which would be the ful-
filment of another promise, and when we have reached sufficient
maturity to think in general terms, we apprehend prima facie
tightness to belong to the nature of any fulfilment of promise.
What comes first in time is the apprehension of the self-
evident prima facie rightness of an individual act of a particular
type. From this we come by reflection to apprehend the self-
evident general principle of prima facie duty. From this, too,
perhaps along with the apprehension of the self-evident prima
j'octef rightness of the same act in virtue of its having another
characteristic as well, and perhaps in spite of the apprehension
of its prima facie wrongness in virtue of its having some third
characteristic, we come to believe something not self-evident
at all, but an object of probable opinion, viz. that this particular
act is (not prima facie but) actually right.
In this respect there is an important difference between
rightness and mathematical properties. A triangle which is
isosceles necessarily has two of its angles equal, whatever other
characteristics the triangle may have—-whatever, for instance,
be its area, or the size of its third angle. The, equality of the
two angles is a parti-resultant attribute.1 And the same is true
of all mathematical attributes. It is true, I may add, of prima
facie rightness. But no act is ever, in virtue of falling under
some general description, necessarily actually right; its right-
ness depends on its whole nature 2 and not on any element in it.
The reason is that no mathematical object (no figure, for instance,
or angle) ever has two characteristics that tend to give it oppo-
site resultant characteristics, while moral acts often (as every one
knows) and indeed always (as on reflection we must admit)
have different characteristics that tend to make them at the
same time prima fade right and prima facie wrong; there is
Cf. pp. *8» IM-J.
* To avoid complicating unduly the statement of the general view I ant puffing
forward, I have here rather overstated it. Any act i$ the origination of a great variety
of things many of which make no difference to its rightness or wrongness. But there
are always many elements in its nature (i. e, in what it is the origination of) that make
a difference to its rightness or wrongness, and no element in its nature can be dismissed
without consideration as indifferent.
probably no act, for instance, which does good to any one
without doing harm to some one else, and. vice versa,

Supposing it to be agreed, as I think on reflection it must,

that no one means by 'right* just 'productive of the best possible
consequences', or 'optitnific*, the attributes 'right* and 'opti-
mific' might stand in either of two kinds of relation to each
other, (i) They might be so related that we could apprehend
a priori^ either immediately or deductively, that any act that
is optimific is right and any act that is right is optimific, as we
can apprehend that any triangle that is equilateral is equiangular
and vice versa. Professor Moore's view is, I think, that the
coextensiveness of 'right' and 'optimific' is apprehended im-
mediately,1 He rejects the possibility of any proof of it. Or
(2) the two attributes might be such that the question whether
they are invariably connected had to be answered by means of
an inductive inquiry. Now at first sight it might seem as if the
constant connexion of the two attributes could be immediately
apprehended. It might seem absurd to suggest that it could be
right for any one to do an act which would produce conse-
quences less good than those which would be produced by
some other act in his power. Yet a little thought will convince
us that this is not absurd. The type of case in which it is easiest
to see that this is so is, perhaps, that in which one has made a
promise. In such a case we all think that prima, facie it is our
duty to fulfil the,promise irrespective of the precise goodness
of the total consequences. And though we do not think it is
necessarily our actual or absolute duty to do so, we are far from
thinking that any, even the slightest, gain in the valpe of the
total consequences will necessarily justify us in doing some-
thing else instead. Suppose, to simplify the case by abstraction,
that the fulfilment of a promise to A would produce 1,000 units
of good 2 for him, but that by doing some other act I could
produce 1,001 units of good for J?, to whom I have made no
Ettics, »8i.
I am assuming that good is objectively quantitative (cf. pp. 142-4), but not that
we can accurately assign an exact quantitative measure to it. Since it is of a definite
amount, we can make the supposition that its amount is so-and-so, though we cannot
with any confidence atsen that it is.
promise, the other consequences of the two acts being of equal
value; should we really think it self-evident that it was our
duty to do the second act and not the first ? I think not. We
should, I fancy, hold that only a much greater disparity of
value between the total consequences would justify us in failing
to discharge our prima facie duty to A* After all, a promise is
a promise, and is not to be treated so lightly as the theory we
are examining would imply. What, exactly, a promise is, is
not so easy to determine, but we are surely agreed that it con-
stitutes a serious moral limitation to our freedom of action.
To produce the 1,001 units of good for B rather than fulfil our
promise to A would be to take, not perhaps our duty as phil-
anthropists too seriously, but certainly our duty as makers of
promises too lightly.
Or consider another phase of the same problem. If I have
promised to confer on A a particular benefit containing 1,000
units of good, is it self-evident that if by doing some different
act I could produce 1,001 units of good for A himself (the
other consequences of the two acts being supposed equal in
value), it would be right for me to do so ? Again, I think not.
Apart from my general prima facie duty to do A what good I
can, I have another prima facie duty to do Mm the particular
service I have promised to do him, and this is not to be set
aside in consequence of a disparity of good of the order of
1,001 to 1,000, though a much greater disparity might justify
me in so doing.
Or again, suppose that A is a very good and J* a very bad
man, should I then, even when I have made no promise, think
it self-evidently right to produce 1,001 units of good for B
rather than 1,000 for At Surely not, I should be sensible of
a prima facie duty of justice, i.e. of producing a distribution of
goods in proportion to merit, which is not outweighed by
such a slight disparity in the total goods to be produced,
Such instances—and they might easily be added to—make
it clear that there is no self-evident connexion between the
attributes 'right' and 'optimific*. The theory we are examining
has a certain attractiveness when applied to our decision that
a particular act is our duty (though I have tried to show that
it does not agree with our actual moral judgements even here),
But it is not even plausible when applied to our recognition
of prima facie duty. For If it were self-evident that the right
coincides with the optimific, it should be self-evident that what
is prima fafie right is prima facie optimific. But whereas we are
certain that keeping a promise is prima facie right, we are
not certain that it is prima facie optimific (though we are per-
hapscertain thatit isprima facie bonific). Our certainty that it is
prima facie right depends not on its consequences but on its
being the fulfilment of a promise. The theory we are examining
involves too much difference between the evident ground of
our conviction about prima facie duty and the alleged ground
of our conviction about actual duty.
The coextensiveness of the right and the optimific is, then,
not self-evident. And I can see no way of proving it deduc-
tively; nor, so far as I know, has any one tried to do so. There
remains the question whether it can be established inductively,
Such an inquiry, to be conclusive, would have to be very
thorough and extensive. We should have to take a large variety
of the acts which we, to the best of our ability, judge to be
right. We should have to trace as far as possible their conse-
quences, not only for the persons directly affected but also for
those indirectly affected, and to these no limit can be set. To
make our inquiry thoroughly conclusive, we should have to do
what we cannot do, viz. trace these consequences into an un-
ending future. And even to make it reasonably conclusive,
we should have to trace them far into the future. It is clear
that the most we could possibly say is that a large variety of
typical acts that are judged right appear, so far as we can trace
their consequences, to produce more good than any other acts
possible to the agents in the circumstances. And such a result
falls far short of proving the constant connexion of the two
attributes. But it is surely clear that no inductive inquiry justi-
fying even this result has ever been carried through. The
advocates of utilitarian systems have been so much persuaded
either of the identity or of the self-evident connexion of the
attributes 'right* and 'optimific' (or 'felicific') that they have
not attempted even such an inductive inquiry as is possible.
And in view of the enormous complexity of the task and the
inevitable inconclusiveness of the result, it is worth no one's
while to make the attempt. What, after all, would be gained
by it? If, as I have tried to show, for an act to be right and to
be optiraific are not the same thing, and an act's being optimific
is not even the ground of its being right, then if we could ask our-
selves (though the question is reallyunmeaning) which we ought
to do, right acts because they are right or optimific acts because
they are optimific, our answer must be *the former'. If they are
optimific as well as right, that is interesting but not morally im-
portant; if not, we still ought to do them (which is only another
way of saying that they are the right acts), and the question
whether they are optimifie has no importance for moral theory.
There is one direction in which a fairly serious attempt has
been made to show the connexion of the attributes 'right* and
'optimific*. One of the most evident facts of our moral con-
sciousness is the sense which we have of the sanctity of promises,
a sense which does not, on the face of it, involve the thought
that one will be bringing more good into existence by fulfilling
the promise than by breaking it. It is plain, I think, that in our
normal thought we consider that the fact that we have made a
promise is in itself sufficient to create a duty of keeping it, the
sense of duty resting on remembrance of the past promise and
not on thoughts of the future consequences of its fulfilment.
Utilitarianism tries to show that this is not so, that the sanctity
of promises rests on the good consequences of the fulfilment of
them and the bad consequences of their non-fulfilment. It
does so in this way: it points out that when you break a promise
you not only fail to confer a certain advantage on your promisee
but you diminish his confidence, and indirectly the confidence
of others, in the fulfilment of promises, You thus strike a blow
at one of the devices that have been found most useful in the
relations between man and man—the device on which, for
example, the whole system of commercial credit rests—and
you tend to bring about a state of things wherein each man,
being entirely unable to rely on the keeping of promises by
others, will have to do everything for himself, to the enormous
impoverishment of human well-being.
To pot the matter otherwise, utilitarians say that when a
promise ought to be kept it is because the total good to be pro-
duced by keeping it is greater than the total good to be
produced by breaking it, the former including as its main ele-
ment the maintenance and strengthening of general mutual
confidence, and the latter being greatly diminished by a weaken-
ing of this confidence. They say, in fact, that the case I put
some pages backl never arises—the case in which by fulfilling
a promise I shall bring into being 1,000 units of good for my
promisee, and by breaking it 1,001 units of good for some one
else, the other effects of the two acts being of equal value. The
other effects, they say, never are of equal value. By keeping
my promise I am helping to strengthen the system of mutual
confidence; by breaking it I am helping to weaken thisj so that
really the first act produces i,ooo+x units of good, and the
second itooi—y units, and the difference between +x and
—y is enough to outweigh the slight superiority in the im-
mediate effects of the second act. In answer to this it may be
pointed out that there must be some amount of good that ex-
ceeds the difference between +x and —y (i.e. exceeds x+y);
say, x+y+ {. Let us suppose the immediate good effects of the
second act to be assessed not at 1,001 but at i,ooo+x+y+^,
Then its net good effects are i,ooo+*+{» i.e. greater than
those of the fulfilment of the promise; and the utilitarian is
bound to say forthwith that the promise should be broken.
Now, we may ask whether that is really the way we think about
promises? Do we really think that the production of the
slightest balance of good, no matter who will enjoy it, by the
breach of a promise frees us from the obligation to keep our
promise? We need not doubt that a system by which promises
are made and kept is one that has great advantages for the
general well-being. But that is not the whole truth. To make
a promise is not merely to adapt an ingenious device for pro-
moting the general well-being; it is to put oneself in a new
relation to one person in particular, a relation which creates
a specifically new prima facie duty to him, not reducible to the
duty of promoting the general well-being of society. By all
« p. 34-
means let us try to foresee the net good effects of keeping one's
promise and the net good effects of breaking it, but even if we
assess the first at 1,000+* and the second at i,ooo+*+{, the
question still remains whether it is not our duty to fulfil the
promise. It may be suspected, too, that the effect of a single
keeping or breaking of a promise in strengthening or weaken-
ing the fabric of mutual confidence is greatly exaggerated by
the theory we are examining. And if we suppose two men
dying together alone, do we think that the duty of one to fulfil
before he dies a promise he has made to the other would be
extinguished by the fact that neither act would have any effect
on the general confidence? Any one who holds this may be
suspected of not having reflected on what a promise is.
I conclude that the attributes 'right* and 'optitnific' are not
identical, and that we do not know either by intuition, by de-
duction, or by induction that they coincide in their application,
still less that the latter is the foundation of the former. It must
be added, however, that if we are ever under no special
obligation such as that of fidelity to a promisee or of gratitude
to a benefactor, we ought to do what will produce most good;
and that even when we are under a special obligation the ten-
dency of acts to promote general good is one of the main factors
in determining whether they are right.

In what has preceded, a good deal of use has been made of

'what we really think* about moral questions; a certain theory
has been rejected because it does not agree with what we really
think. It might be said that this is in principle wrong; that we
should not be content to expound what our present moral
consciousness tells us but should aim at a criticism of our exist-
ing moral consciousness in the light of theory. Now I do not
doubt that the moral consciousness of men has in detail under-
gone a good deal of modification as regards the things we think
right, at the hands of moral theory. But if we are told, for
instance, that we should give up our view that there is a special
obligatoriness attaching to the keeping of promises because it
is self-evident that the only duty is to produce as much good as
possible, we have to ask ourselves whether we really, when we
reflect, are convinced that this is self-evident, and whether we
really can get rid of our view that promise-keeping has a bind-
ingness independent of productiveness of maximum good. In
my own experience I find that I cannot, in spite of a very genuine
attempt to do so; and I venture to think that most people will
find the same, and that just because they cannot lose the sense
of special obligation, they cannot accept as self-evident, or even
as true, the theory which would require them to do so. In fact
it seems, on reflection, self-evident that a promise, simply as
such, is something that prima facie ought to be kept, and it
does not, on reflection, seem self-evident that production of
maximum good is the only thing that makes an act obligatory.
And to ask us to give up at the bidding of a theory our actual
apprehension of what is right and what is wrong seems like
asking people to repudiate their actual experience of beauty,
at the bidding of a theory which says 'only that which satisfies
such and such conditions can be beautiful'. If what I have
called our actual apprehension is (as I would maintain that it
is) truly an apprehension, i.e. an instance of knowledge, the
request is nothing less than absurd.
I would maintain, in fact, that what we are apt to describe
as "what we think* about moral questions contains a consider-
able amount that we do not think but know, and that this forms
the standard by reference to which the truth of any moral
theory has to be tested, instead of having itself to be tested by
reference to any theory. I hope that I have in what precedes
indicated what in my view these elements of knowledge are
that are involved in our ordinary moral consciousness.
It would be a mistake to found a natural science on 'what we
really think', i.e. on what reasonably thoughtful and well-
educated people think about the subjects of the science before
they have studied them scientifically. For such opinions are
interpretations, and often misinterpretations, of sense-ex-
perience; and the man of science must appeal from these to
sense-experience itself, which furnishes his real data. In ethics
no such appeal is possible. We have no more direct way of
access to the facts about Tightness and goodness and about
what things are right or good, than by thinking about them;
the moral convictions of thoughtful and well-educated people
are the data of ethics just as sense-perceptions are the data of a
natural science. Just as some of the latter have to be rejected
as illusory, so have some of the former; but as the latter are
rejected only when they are in conflict with other more accurate
sense-perceptions, the former are rejected only when they are
in conflict with other convictions which stand better the test of
reflection. The existing body of moral convictions of the best
people is the cumulative product of the moral reflection of many
generations, which has developed an extremely delicate power
of appreciation of moral distinctions; and this the theorist
cannot afford to treat with anything other than the greatest
respect. The verdicts of the moral consciousness of the best
people are the foundation on which he must build; though he
must first compare them with one another and eliminate any
contradictions they may contain.

It is worth while to try to state more definitely the nature of

the acts that are right. We may try to state first what (if any-
thing) is the universal nature of all acts that are right. It is
obvious that any of the acts that we do has countless effects,
directly or indirectly, on countless people, and the probability
is that any act, however right it be, will have adverse effects
(though these may be very trivial) on some innocent people.
Similarly, any wrong act will probably have beneficial effects
on some deserving people. Every act therefore, viewed in
some aspects, will be prima facie right, and viewed in others,
prim® facie wrong, and right acts can be distinguished from
wrong acts only as being those which, of all those possible for
the agent in the circumstances, have the greatest balance of
prima facie rightness, in those respects in which they ate prima
facie right, over their prims fade wrongness, in those respects
in which they are prima facie wrong—prima facie rightness
and wrongness being understood in the sense previously ex-
plained. For the estimation of the comparative stringency of
these prima facie obligations no general rules can, so far as I can
see, be laid down. We can only say that a great deal of strin-
gency belongs to the dudes of 'perfect obligation'—the dudes
of keeping our promises, of repairing wrongs we have done, and
of returning the equivalent of services we have received. For
the rest, ev rjj aiaBrpei if Kptms," This-sense of our particular
duty in particular circumstances, preceded and informed by
the fullest reflection we can bestow on the act in all its bearings,
is highly fallible, but it is the only guide we have to our duty.
"When we turn to consider the nature of individual right
acts, the first point to which attention should be called is that
any act may be correctly described in an indefinite, and in
principle infinite, number of ways. An act is the production
of a change in the state of affairs (if we ignore, for simplicity's
sake, the comparatively few cases in which it is the maintenance
of an existing state of affairs; cases which, I think, raise no
special difficulty). Now the only changes we can directly pro-
duce are changes in our own bodies or in our own minds. But
these are not, as such, what as a rule we think it our duty to
produce. Consider some comparatively simple act, such as
telling the truth or fulfilling a promise. In the first case what I
produce directly is movements of my vocal organs. But what
I think it my duty to produce is a true view in some one else's
mind about some fact, and between my movement of my vocal
organs and this result there intervenes a series of physical events
and events in his mind. Again, in the second case, I may have
promised, for instance, to return a book to a friend. I may be
able, by a series of movements of my legs and hands, to place
it in his hands. But what I am just as likely to do, and to think
I have done my duty in doing, is to send it by a messenger or
to hand it to his servant or to send it by post; and in each of
these cases what I do directly is worthless in itself and is con-
nected by a series of intermediate links with what I do think it
is my duty to bring about, viz, his receiving what I have
promised to return to him. This being so, it seems as if what
I do has no oblSgatoriness in itself and as if one or other of three
accounts should be given of the matter, each of which makes
Tightness not belong to what I do, considered in its own nature.
(i) One of them would be that what is obligatory is not
doing anything in the natural sense of producing any change
'The decision rests with perception*. Arist, Nk. EtL 1109 b ij,»(2(5 b 4.
in the state of affairs, but aiming at something—at, for instance,
my friend's reception of the book. But this account will not do.
For (a) to aim at something is to act from a motive consisting
of the wish to bring that thing about. But we have seen1 that
motive never forms part of the content of our duty; if anything
is certain about morals, that, I think, is certain. And ($) if I
have promised to return the book to my friend, I obviously do
not fulfil my promise and do my duty merely by aiming at his
receiving the book; I must see that he actually receives it.
(2) A more plausible account is that which says I must do that
which is likely to produce the result. But this account is open
to the second of these objections, and probably also to the first.
For in the first place, however likely my act may seem, even on
careful consideration, and even however likely it may in fact be,
to produce the result, if it does not produce it I have not done
what I promised to do, i.e. have not done my duty. And
secondly, when it is said that I ought to do what is likely to
produce the result, what is probably meant is that I ought to do
a certain thing as a result of the wish to produce a certain result,
and of the thought that my act is likely to produce it; and this
again introduces motive into the content of duty. (3) Much the
most plausible of the three accounts is that which says, *I ought
to do that which will actually produce a certain result.' This
escapes objection ($). Whether it escapes objection (a) or not
depends on what exactly is meant. If it is meant that I ought
to do a certain thing from the wish to produce a certain result
and the thought that it will do so, the account is still open to
objection (a). But if it is meant simply that I ought to do a
certain thing, and that the reason why I ought to do it is that it
will produce a certain result, objection (a) is avoided. Now
this account in its second form is that which utilitarianism
gives. It says what is right is certain acts, not certain acts
motivated in a certain way; and it says that acts are never right
by their own nature but by virtue of the goodness of their
actual results. And this account is, I think, clearly nearer the
truth than one which makes the rightness of an act depend on
the goodness of either the intended'or the likely results.
pp. f-6.
Nevertheless, this account appears not to be the true one.
For it implies that what we consider right or our duty is what
we do directly. It is this, e.g. the packing up and posting of
the book, that derives its moral significance not from its own
nature but from its consequences. But this is not what we
should describe, strictly, as our duty; our duty is to fulfil our
promise, i.e. to put the book into our friend's possession. This
we consider obligatory in its own nature, just because it is a
fulfilment of promise, and not because of its consequences.
But, it might be replied by the utilitarian, I do not do this; I
only do something that leads up to this, and what I do has no
moral significance in itself but only because of its consequences.
In answer to this, however, we may point out that a cause
produces not only its immediate, but also its remote conse-
quences, and the latter no less than the former. I, therefore,
not only produce the immediate movements of parts of my body
but also my friend's reception of the book, which results from
these. Or, if this be objected to on the grounds that I can hardly
be said to have produced my friend's reception of the book
when I have packed and posted it, owing to the time that has
still to elapse before he receives it, and that to say I have pro-
duced the result hardly does justice to the part played by the
Post Office, we may at least say that I have secured my friend's
reception of the book. What I do is as truly describable in this
way as by saying that it is the packing and posting of a book.
(It is equally truly describable in many other ways; e.g. I have
provided a few moments' employment for Post Office officials.
But this is irrelevant to the argument.) And if we ask ourselves
whether it is qua, the packing and posting of a book, or qua the
securing of my friend's getting what I have promised to return
to him, that my action is right, it is clear that it is in the second
capacity that it is right; and in this capacity, the only capacity
in which it is right, it is right by its own nature and not because
of its consequences.
This account may no doubt be objected to, on the ground
that we are ignoring the freedom of will of the other agents-—
the sorter and the postman, for instance—who are equally
responsible for the result. Society, it may be said, is not like a
machine, in which event follows event by rigorous necessity.
Some one may, for instance, in the"exercise of his freedom of
will, steal the book on the way. But it is to be observed that I
have excluded that case, and any similar case. I am dealing with
the case in which I secure my friend's receiving the book;, and
if he does not receive it I have not secured his receiving it. If
on the other hand the book reaches its destination, that alone
shows that, the system of things being what it is, the trains by
which the book travels and the railway lines along which it
travels being such as they are and subject to the laws they are
subject to, the postal officials who handle it being such as they
are, having the motives they have and being subject to the
psychological laws they are subject to, my posting the book
was the one further thing which was sufficient to procure my
friend's receiving it If it had not been sufficient, the result
would not have followed. The attainment of the result proves
the sufficiency of the means. The objection in fact rests on the
supposition that there can be unmotived action, i.e. an event
without a cause, and may be refuted by reflection on the uni-
versality of the law of causation,
It is equally true that non-attainment of the result proves
the insufficiency of the means. If the book had been destroyed
in a railway accident or stolen by a dishonest postman, that
would prove that my immediate act was not sufficient to pro-
duce the desired result. We get the curious consequence that
however carelessly I pack or dispatch the book, if it comes to
hand I have done my duty, and however carefully I have acted,
if the book does not come to hand I have not done my duty.
Success and failure are the only test, and a sufficient test, of the
performance of duty. Of course, I should deserve more praise
in the second case than in the first; but that is an entirely different
question; we must not mix up the question of right and wrong
with that of the morally good and the morally bad. And that
our conclusion is not as strange as at first sight it might seem
is shown by the fact that if the carelessly dispatched book
comes to hand, it is not my duty to send another copy, while
if the carefully dispatched book does not come to hand I must
send another copy to replace it. In the first case I have not my
duty still to do, which shows that I have done it; in the second
I have it still to do, which shows that I have not done it.
We have reached the result that my act is right fw being an
ensuring of one of the particular states of affairs of which it is
an ensuring, viz., in the case we have taken, of my friend's
receiving the book I have promised to return to him. But this
answer requires some correction; for it refers only to the prima
facie rightness of my act. If to be a fulfilment of promise were
a sufficient ground of the rightness of an act, alt fulfilments of
promises would be right, whereas it seems clear that there are
cases in which some othet prima facig duty overrides the prima
facie duty of fulfilling a promise. The more correct answer
would be that the ground of the actual rightness of the act is
that, of all acts possible for the agent in the circumstances, it
is that whose prima facie rightness in the respects in which it is
prima facie right most outweighs its prima fade wrongness in
any respects in which it is prima facie wrong. But since its
prima Jade rightness is mainly due to its being a fulfilment of
promise, we may call its being so the salient element in the
ground of its rightness.
Subject to this qualification, then, it is as being the produc-
tion (or if we prefer the word, the securing or ensuring) of the
reception by my friend of what I have promised him (or in
other words as the fulfilment of my promise) that my act is
right It is not right as a packing and posting of a book. The
packing and posting of the book is only incidentally right,
right only because it is a fulfilment of promise, which is what is
directly or essentially right.
Our duty, then, is not to do certain things which will pro-
duce certain results. Our acts, at any rate out acts of special
obligation, are not right because they will produce certain
results—which is the view common to all forms of utilita-
rianism. To say that is to say that in the case in question what
is essentially right is to pack and post a book, whereas what is
essentially right is to secure the possession by my friend of
what I have promised to return to him. An act is not right
because it, being one thing, produces good results different from
itself; it is right because it is itself the production of a certain
state of affairs. Such production is right in itself, apart from
any consequence,
But, it might be said, this analysis applies only to acts of
special obligation; the utilitarian account still holds good for
the acts in which we are not under a special obligation to any
person or set of persons but only under that of augmenting the
general good. Now merely to have established that there are
special obligations to do certain things irrespective of their con-
sequences would be already to have made a considerable breach
in the utilitarian walls; for according to utilitarianism there is
no such thing, there is only the single obligation to promote
the general good. But, further, on reflection it is clear that just
as (in the case we have taken) my act is not only the packing
and posting of a book but the fulfilling of a promise, and just
as it is in the latter capacity and not in the former that it is my
duty, so an act whereby I augment the general good is not only,
let us say, the writing of a begging letter on behalf of a hospital,
but the producing (or ensuring) of whatever good ensues
therefrom, and it is in the latter capacity and not in the former
that it is right, if it w right. That which is right is right not
because it is an act, one thing, which will produce another
thing, an increase of the general welfare, but because it is itself
the producing of an increase in the general welfare. Or, to
qualify this in the necessary way, its being the production of an
increase in the general welfare is the salient element in the
ground of its Tightness. Just as before we were led to recognize
the prima facie lightness of the fulfilment of promises, we are
now led to recognize the prima facie rightness of promoting
the general welfare. In both cases we have to recognize the
intrinsic rightness of a certain type of act, not depending on its
consequences but on its own nature.

A general discussion of right or duty would hardly be com-
plete without some discussion, even if only a brief one, of the
closely related subject of rights. It is commonly said that
rights and duties are correlative, and it is worth while to inquire
whether and, if at all, in what sense this is true. The statement
may stand for any one, or any combination, of the following
logically independent statements:
(1) A right of A against B implies a duty of B to A,
(2) A duty of S to ^implies a right of A against B,
(3) A right of A against B implies a duty of A to B.
(4) A duty of A to B implies a right of A against B.
What is asserted in (i) is that A's, having a right to have a
certain individual act done to him by B implies a duty for B
to do that act to A; (2) asserts the converse implication; what
is meant by (3) is that A's having a right to have a certain act
done to him by B implies a duty for A to do another act to B,
which act may be either a similar act (as where the right of
having the truth told to one implies the duty of telling the
truth) or a different sort of act (as where the right to obedience
implies the duty of governing well); (4) asserts the converse
Of these four propositions the first appears to be unquestion-
ably true; a right in one being against another is a right to treat
or be treated by that other in a certain way, and this plainly
implies a duty for the other to behave in a certain way. But there
is a certain consideration which throws doubt on the other three
propositions. This arises from the fact that we have duties to
animals and to infants. The latter case is complicated by the
fact that infants, while they are not (so we commonly believe)
actual moral agents, are potential moral agents, so that the duty
of parents, for instance, to support them may be said to be
counterbalanced by a duty which is not incumbent on the
infants at the time but will be incumbent on them later, to obey
and care for their parents. We had better therefore take the
less complicated case of animals, which we commonly suppose
not to be even potential moral agents.
It may of course be denied that we have duties to animals,
The view held by some writers is that we have duties concern-
ing animals but not to them, the theory being that we have a
duty to behave humanely to our fellow men, and that we should
behave humanely to animals simply for fear of creating a dis-
position in ourselves which will make us tend to be cruel to
our fellow men. Professor D, G, Ritchie, for instance, implies
that we have not a duty to animals except in a sense like that
in which the owner of an historic house may be said to have a
duty to the house,' Now the latter sense is, I suppose, purely
metaphorical We may in a fanciful mood think of a noble
house as if it were a conscious being having feelings which we
are bound to respect. But we do not really think that it has them,
I suppose that the duty of the owner of an historic house is
essentially a duty to his contemporaries and to posterity; and
he may also think it is a duty to his ancestors. On die other
hand, if we think we ought to behave in a certain way to ani-
mals, it is out of consideration primarily for their feelings that
we think we ought to behave so; we do not think of them
merely as a practising-ground for virtue. It is because we think
their pain a bad thing that we think we should not gratuitously
cause it. And I suppose that to say we have a duty to so-and-so
is the same thing as to say that we have a duty, grounded on
facts relating to them, to behave in a certain way towards them,
Now if we have a duty to animals, and they have not a duty
to us (which seems clear, since they are not moral agents), the
first and last of our four propositions cannot both be true, since
(4) implies that a duty of men to animals involves a right of
men against animals, and (i) implies that this involves a duty
of animals to men, and therefore (4) and (i) together imply
that a duty of men to animals involves a duty of animals to men.
And since the first proposition is clearly true, the fourth must
be false; it cannot be true that a duty of A to B necessarily
involves a right of ^against B, Similarly, the second and third
propositions cannot both be true; for (2) and (3) taken together
Natural Rights, 108.
imply that a duty of men to animals involves a duty of animals
to men. But here it is not so clear which of the two propositions
is true; for it is not clear whether we should say that though we
have a duty to animals they have no right against us, or that
though they have a right against us they have no duty to us.
If we take the first view, we are implying that in order to have
rights, just as much a$ in order to have duties, it is necessary to
be a moral agent. If we take the second view, we are implying
that while only moral agents have duties, the possession of a
nature capable of feeling pleasure and pain is all that is needed
in order to have rights. It is not at all clear which is the true
view. On the whole, since we mean by a right something that
can be justly claimed, we should probably say that animals
have not rights, not because the claim to humane treatment
would not be just if it were made, but because they cannot
make it. But the doubt which we here find about the applica-
tion of the term 'rights* is characteristic of the term. There are
other ways too in which its application is doubtful. Even if we
hold that it is our duty not merely to do what is just to others
but to promote their welfare beyond what justice requires, it
is not at all clear that we should say they have a right to benefi-
cent treatment over and above what is just. We have a tendency
to think that not every duty incumbent on one person involves
a right in another.
This characteristic of our way of thinking about rights has
been fastened upon by theory. Green, for instance, divides the
whole region of duty into three parts: (i) moral duties which
involve no rights on the other side, (2) obligations involving
such rights, both obligations and rights being included in the
jus naturae and being such as should be legally recognized,
(3) legal obligations involving legal rights on the other side.'
He describes the rights in class (2)—what I will for brevity call
moral rights—as sharing with legal rights the characteristic of
depending for their existence on some form of general recogni-
tion. The recognition in the latter case consists in the making
of a lawj in the former it consists simply in a general state of
public opinion. Now it is plainly wrong to describe either
Priacipltt of Political OUigatian, §§ to, 11.
legal or moral rights as depending for their existence on their
recognition, for to recognize a thing (in the sense in which
'recognize' is here used) is to recognize it as existing already.
The promulgation of a law is not the recognition of a legal
right, but the creation of it, though it may imply the recognition
of an already existing moral right. And to make the existence
of a moral right depend on its being recognized is equally mis-
taken. It would imply that slaves, for instance, acquired the
moral right to be free onlyatthemomentwhenamajority of man-
kind, or of some particular community, formed the opinion that
they ought to be free, i.e. when the particular person whose con-
version to this view changed a minority into a majority changed
his mind. Such a view, of course, cannot be consistently main-
tained, and we find Green implying in successive sections that
social recognition is indispensable to the existence of rights,1
and that the slave has a right to citizenship though this right is
not recognized by society.2 In the latter passage we see the
true Green, the passionate lover of liberty, reacting against the
theory of the previous page. Some may think that slavery
is not wrong; but every one will admit that there are certain
forms of treatment of others which are wrong and which the
sufferer has the right to have removed, whether this right is
recognized by society or not.
There is, however, to be found in Green another view which
is less clearly false. According to this, the existence of a right
is made to depend not on the recognition of it but on the recog-
nition of a power in the person in question to seek an end com-
mon to all the citizens of a community.3 This avoids the patent
error of making the existence of a right depend on its being
recognized to exist. Yet like the former view it makes a moral
right depend not on the nature of a given person and his rela-
tions to his fellows, but on what people think about them, i.e.
on what a majority of the community think about them. But
'A claim to which reality is given by social recognition, and thus implicitly a
right* {§ 139). Cf. 'This recognition of a power, ifi some way or other, as dwt which
should be, is always necessary to render it a right* (§23). 'Rights are made by recogni-
tion. There is no right "but thinking makes it so" ' {§ 136).
§ 140 implies that the slave's right to citizenship is founded on his possessing a
common human consciousness with the citizens of the slate,
J Cf. e. g. §§ 25, 16.
though the existence of legal rights depends on the degree of
enlightenment of the community, the existence of moral rights
plainly does not, but on the nature and relations of the persons
Green's theory seems to have arisen as follows. He starts
his historical survey with Hobbes and Spinoza, both of whom
identify right with power. A legal right may be identified with
a certain kind of power; it is the power of getting certain things
not by one's own brute force but by the aid of the law. Green
seems to have tried to get a theory of moral rights by making
a similar amendment of the bare identification of right with
power; and he accordingly identifies them with the power of
getting certain things not by one's own brute force nor by the
aid of the law but by the aid of public opinion; instead of saying,
what is surely evident, that a moral right is not a power at all.
Yet there are elements in his account which point to a truer
theory; e.g. *a "right" is an ideal attribution ("ideal" in the sense
of not being sensibly verifiable)*,* Now whether a given society
recognizes a particular right is, I take it, sensibly verifiable in
the sense in which Green here insists that a right is not. What
is not sensibly verifiable is whether the society is justified in
recognizing the right, and this depends on whether the right
is there antecedently to society's recognition of it. Thus the
insistence that a right is not sensibly verifiable points to an
objective theory of rights; but unfortunately Green follows
this clue no farther.
If we eliminate the possibility of holding that animals have
rights, by saying that only that which has a moral nature can have
a right, our main doubt with regard to the correlation of rights
and duties is on the question whether there is a right to benefi-
cence. It is obvious that a man has a right to just treatment, and
it is commonly agreed that he has a right to have promises made
to him fulfilled; it is less generally agreed that he has a right to
beneficent treatment, even when it is admitted that it is our
duty to treat him beneficently.
Some would even say that to treat others beneficently is to
go beyond our duty. But probably this statement rests on a
mere confusion. We usually oppose justice to benevolence,
But while treating a man justly is,commonly understood to
mean doing certain things to him (paying our debts to him,
and the like), irrespective of the spirit in which we do them,
treating him benevolently obviously means doing certain
things to him from goodwill. And it is rightly felt that there
is a great difference between the two things, and it is found
natural to say that the one implies, and the other does not, a
right on the other side, and (by some people) even to say that
the one is a duty and the other is not. But if we will distinguish
between doing what is just and doing it in the spirit of justice,
and between doing what is beneficent and doing it in the spirit
of beneficence, then (in accordance with the principle that it is
always acts, and not acts from a certain motive, that are our
duty) it is clear that it is not our duty to act in the spirit of
justice, any more than in the spirit of beneficence, and that it
Is our duty to do what is beneficent, as it is our duty to do what
is just,
If we are clear on this point, our main objection to saying
that the other person has a right to beneficence disappears, I
do not say that our whole objection disappears; for there hangs
about the notion of a "right* the notion of its being not only
something which one person should in decency respect but
also something which the other person can in decency claim,
and we feel that there is something indecent in the making of a
clam to beneficence.
These doubts about the application of the term 'right* appear
to spring from the fact that 'right* (the noun) does not stand
for a purely moral notion. It began, I suppose, by standing for
a legal notion, and its usage has broadened out so as to include
certain things that cannot be claimed at law; but its usage has
not yet broadened out so much as to become completely cor-
relative to duty. Once we start on the process of broadening
it out, however, there seems to be no secure resting-place short
of this.
Returning now to the four propositions about the correla-
tivity of duties and rights, it seems that with regard to the
second proposition, *A duty of B to A implies a right of A
against £' (which has latterly been the subject of our discussion),
we should say (i) that this is not true when A is not a moral
agent, and (2) that it is true when A is a moral agent (even if
the duty be the duty of beneficent action). And since our only
doubt about the third proposition, *A right of A against B
implies a duty of A to B\ arises from our doubt whether ani-
mals have not rights, if we agree that animals have not rights
we need not doubt the truth of this proposition. It is this pro-
position, above all, that has been maintained by those who have
insisted on the correlativity of rights and duties; for this was
maintained essentially against the belief that men have 'natural
rights' in a state of nature in which they have no duties.
A further problem, however, awaits us, viz. whether a failure
to do one's duty involves a corresponding loss of right. Or
rather, as we have found the meaning of "rights' more doubtful
than that of 'duties', it will be more profitable to omit any
reference to rights, and put our question in the form, 'if A fails
in his duty to B, does that put an end to £'$ duty to A? In
some cases we seem to be clear that this is so. If a tradesman
sends me goods inferior to those I chose in his shop, I am not
morally, any more than legally, bound to pay him the full
price; I may return the goods and pay nothing, or (with his
consent) keep them and pay a lower price. And in general any
duty arising out of a contract is cancelled by non-fulfilment of
the corresponding duty on the other side. In other cases we
are not so clear. It is not so generally agreed, for instance,, that
if A tells lies to j?, B is justified in telling lies to A. Two blacks,
we say in such a case, do not make a white. Yet the peculiar
stringency of the duty of veracity seems to spring from an
implicit understanding that language shall be used to convey
the real opinions of the speakers, and it would seem that a
failure to carry out the understanding on one side makes it no
longer binding on the other; and we should have small patience
with an habitual liar who insisted on strict veracity in others.
It must be admitted that a man who has deceived me has de-
stroyed what would have been the main reason for its being my
duty to tell him the truth. But we should probably hesitate to
say that by his breach of the implicit understanding my duty
to tell him the truth has been entirely destroyed, as by the
tradesman's breach of contract my duty to pay him has been
destroyed. Various reasons help to account for this. For one
thing, it is likely that by deceiving a liar I may indirectly deceive
innocent people; for another, the consequences for my own
character are likely to be particularly dangerous. But the main
reason probably lies elsewhere. Before the contract was made
between my tradesman and me, there was no duty incumbent
on me of paying him this sura of money, I had a general duty
to promote the good of all men, but there was no obvious
reason for supposing that this could be best done by trans-
ferring this sum of money to him. But even before the implicit
undertaking to tell the truth was established I had a duty not to
tell lies, since to tell lies is prima fade to do a positive injury to
another person. Since this duty does not rest on contract, it is
not abolished by the breach of contract, and therefore while a
person who has been deceived by another is justified in refusing
to answer his questions, he is not justified in telling him lies,
Yet that this forms only a small part of the stringency of the
duty of truthfulness may be inferred from the leniency with
which we should judge deceit, in a case in which no implicit
undertaking to tell the truth has been established, e.g. when a
civilized man deceives a savage whom he has just met for the
first time, or vice versat or when one of two savages belonging
to different tribes deceives the other. Deceit is much more
venial in such a case, because the offender has no reason to
suppose that the other is not deceiving, or going to deceive,
Taking, then, the obvious division between duties arising
out of contract and those that arise otherwise, we must say
that while the former are cancelled by breach of the contract
on the other side, the latter are not cancelled by the bad be-
haviour of the other person. It would also seem, from a con-
sideration of our actual moral judgements, that the former
type of duty is the more stringent of the two.
Now the distinction between the rights corresponding to
duties that arise' out of contract, and the rights corresponding
to other duties, may be quite suitably expressed as a distinction
between contractual and natural rights, and the notion of
natural rights as a distinct class may thus be vindicated, if it be
cut free from the belief which has been so often bound up with
it, that there are rights in a state of nature, i.e. in a state in which
there are no duties. Such a belief is made possible for Hobbes
only by a complete confusion between rights and powers,
amounting to an express identification of the two.

In connexion with the discussion of rights it is proper to
consider a question which has always interested and usually
puzzled moralists, and which forms a crucial example for the
testing of moral theories—the question of punishment. A
utilitarian theory, whether of the hedonistic or of the "ideal*
kind, if it justifies punishment at all, is bound to justify it solely
on the ground of the effects it produces. The suffering of pain
by the person who is punished is thought to be in itself a bad
ching, and the bringing of this bad thing into the world is held
to need justification, and to receive it only from the fact that
the effects are likely to be so much better than those that would
follow from his non-punishment as to outweigh the evil of his
pain. The effects usually pointed to are those of deterrence and
of reformation. In principle, then, the punishment of a guilty
person is treated by utilitarians as not different in kind from
the imposition of inconvenience, say by quarantine regulations,
on innocent individuals for the good of the community. Or
again, if a state found to be prevalent some injury to itself or to
its members that had not been legislated against, and proceeded
to punish the offenders, its action would in principle be justified
by utilitarians in the same way as its punishment of offenders
against the law is justified by them, viz. by the good of the
community. No doubt the state would have greater difficulty
in justifying its action, for such action would produce bad
consequences which the punishment of law-breakers does not.
But the difference would be only in degree. Nay more, a govern-
pent which found some offence against the law prevalent, and
in its inability to find the offenders punished innocent people
on the strength of manufactured evidence, would still be able
to justify its action on the same general principle as before.
Plain men, and even perhaps most people who have reflected
on moral questions, are likely to revolt against a theory which
involves such consequences, and to exclaim, that there is all the
difference in the world between such action and the punish-
ment of offenders against the law. They feel the injustice of
such action by the state, and are ready to say, in the words
imputed to them by Mr, Bradley; 'Punishment is punishment,
only when it is deserved. We pay the penalty because we owe
it, and for no other reason; and if punishment is inflicted for
any other reason whatever than because it is merited by wrong,
it is a gross immorality, a crying injustice, an abominable crime,
and not what it pretends to be. We may have regard for what-
ever considerations we please—our own convenience, the good
of society, the benefit of the offender; we are fools, and worse,
if we fail to do so. Having once the right to punish, we may
modify the punishment according to the useful and the pleasant;
but these are external to the matter, they cannot give us a right
to punish, and nothing can do that but criminal desert,*J
There is one form of utilitarian view which differs in an
important respect from that above ascribed to utilitarians.
Professor Moore admits the possibility, which follows from
his doctrine of organic unities, that punishment may not need
to be justified merely by its after-effects. He points out 2 that
it may well be the case that though crime is one bad thing and
pain another, the union of the two in the same person may be
a less evil than crime unpunished, and might even be a positive
good. And to this extent, while remaining perfectly consistent
with his own type of utilitarianism, he joins hands with intui-
tionists, most of whom, at any rate, would probably hold that
the combination of crime and punishment is a lesser evil than
unpunished crime.
Most intuitionists would perhaps take the view that there is
a fundamental and underivative duty to reward the virtuous
Ethical Studies, ed. 2,16-7. * Principle Etkka, 214.
and to punish the vicious. I am inclined to diverge from this
view. Two things seem to me to be clear: that we have a prima
facie duty to do this, and that a state of affairs in which the good
are happy and the bad unhappy is better than one in which the
good are unhappy and the bad happy. Now if the first of these
is an underivative fact, the two facts are logically unconnected.
For it can be an underivative fact only if the intuitionist view
is true, and if that view is true the superiority of the one state
of affairs over the other cannot follow from the duty of pro-
ducing it, since on the intuitionist view there are duties other
than the duty of producing good. But an intuitionist may with
propriety perform the reverse derivation; he may derive the
primafade duty of reward and punishment from the superiority
of the state of affairs produced, since he may—and, as I think,
must—admit that if a state of affairs is better than its alter-
natives there is a prima facie duty to produce it if we can.
The duty of reward and punishment seems to me to be in this
way derivative. It can be subsumed under the duty of pro-
ducing as much good as we can; though it must be remembered
that the good to be produced in this case is very different from
the other goods we recognize (say virtue, knowledge, and
pleasure)} consisting as it does in a certain relative arrange-
ment of virtue, vice, pleasure, and pain.
But if we hold that there is this duty, it must be admitted
that it is one which it is very difficult for us to see our way
to performing, since we know so little about the degrees of
virtue and vice, and of happiness and unhappiness, as they
occur in our fellow men. And in particular there are two grave
objections to holding that the principle of punishing the vicious,
for the sake of doing so, is that on which the state should pro-
ceed in its bestowal of punishments.
(i) What we perceive to be good is a condition of things in
which the total pleasure enjoyed by each person in his life as a
whole is proportional to his virtue similarly taken as a whole.
Now it is by no means clear that we should help to bring about
this end by punishing particular offences in proportion to their
moral badness. Any attempt to bring about such a state of
affairs should take account of the whole character of the persons
involved, as manifested in their life taken as a whole, and of
the happiness enjoyed by them throughout their life taken as
a whole, and it should similarly take account of the virtue taken
as a whole, and of the happiness taken as a whole, of each of
the other members of the community, and should seek to bring
about the required adjustments. In the absence of such a view
of the whole facts, the criminals that a retributive theory of
state punishment would call on us to punish for the sake
of doing so may well be persons who are more sinned against
than sinning, and may be, quite apart from our intervention,
already enjoying less happiness than a perfectly fair distribution
would allow them. The offences which the state legislates
against are only a small part of the wrong acts which are being
done every day, and a system which punishes not all wrong
acts, but only those which have been forbidden by law, and
does not attempt to reward all good acts—such an occasional
and almost haphazard system of intervention does not hold
out any good hope of promoting the perfect proportionment
of happiness to virtue. Nor would it be in the least practicable
for the state to attempt the thorough review of the merit and
the happiness of all its members, which alone would afford a
good hope of securing this end.
(2) Even if it were practicable, it is by no means clear that
it is the business of the state to aim at this end. Such a view
belongs, I think, to an outworn view of the state, one which
identifies the state with the whole organization of the com-
munity. In contrast to this, we have come to look upon the
state as the organization of the community for a particular
purpose, that of the protection of the most important rights of
individuals, those without which a reasonably secure and com-
fortable life is impossible; and to leave the promotion of other
good ends to the efforts of individuals and of other organiza-
tions, such as churches, trade unions, learned and artistic
societies, clubs. Now it cannot, I think, be maintained that the
apportionment of happiness to merit is one of the essential
conditions to the living of a reasonably secure and comfortable
life. Life has gone on for centuries being lived with reasonable
security and comfort though states have never achieved or
even attempted with any degree of resolution to effect this
apportionment. And in fact for the state to make such an
attempt would seriously interfere with its discharge of its
proper work. Its proper work is that of protecting rights.
Now rights are (as we have seen) rights to be treated in certain
ways and not to be treated in certain ways from certain motives;
what the state has to take account of, therefore, is not morally
bad actions, but wrong acts, and it has to take account of them in
such a way as to diminish the chance of their repetition. And
this attempt would only be interfered with if the state were at
the same time trying to effect a proportionment of happiness
to moral worth in its members. The latter task, involving as it
would a complete review of the merit and happiness of all its
members, would involve leaving the punishment for each
offence undetermined by law, and to be determined in the light
of all the circumstances of each case; and punishment so com-
pletely undetermined in advance would be quite ineffective as
a protector of rights.
But to hold that the state has no duty of retributive punish-
ment is not necessarily to adopt a utilitarian view of punishment.
It seems possible to give an account of the matter which retains
elements in punishment other than that of expediency, without
asserting that the state has any duty properly defined as the duty
of punishing moral guilt. The essential duty of the state is to
protect the most fundamental rights of individuals. Now,
rights of any human being are correlative to duties incumbent
on the owner of rights, or, to put it otherwise, to rights owned
by those against whom he has rights; and the main element in
any one's right to life or liberty or property is extinguished by
his failure to respect the corresponding right in others.1 There
is thus a distinction in kind which we all in fact recognize, but
which utilitarianism cannot admit, between the punishment of
a person who has invaded the rights of others and the infliction
of pain or restraint on one who has not. The state ought, in its
effort to maintain the rights of innocent persons, to take what
steps are necessary to prevent violations of these rights; and
the offender, by violating the life or liberty or property of
• a. pp. 54-5.
another, has lost his own right to have his life, liberty, or
property respected, so that the state has no prima facie duty
to spare him, as it has a prima. facie duty to spare the innocent,
It is morally at liberty to injure him as he has injured others,
or to inflict any lesser injury on him, or to spare him, exactly as
consideration both of the good of the community and of his
own good requires. If, on the other hand, a man has respected
the rights of others, there is a strong and distinctive objection
to the state's inflicting any penalty on him with a view to the
good of the community or even to his own good. The interests
of the society may sometimes be so deeply involved as to make it
right to punish an innocent man 'that the whole nation perish
not*. But then the prima facie duty of consulting the general
interest has proved more obligatory than the perfectly distinct
prima facie duty of respecting the rights of those who have
respected the rights of others.
This is, I believe, how most thoughtful people feel about
the affixing of penalties to the invasion of the rights of others.
They may have lost any sense they or their ancestors had that
the state should inflict retributive punishment for the sake of
doing so, but they feel that there is nevertheless a difference
of kind between the community's right to punish people for
offences against others, and any right it may have to incon-
venience or injure innocent people in the public interest. This
arises simply from the fact that the state has A prima facie duty
not to do the latter and no such duty not to do the former,
We can, I think, help ourselves towards an understanding of
the problem by distinguishing two stages which are not usually
kept apart in discussions of it. The infliction of punishment by
the state does not, or should not, come like a bolt from the blue,
It is preceded by the making of a law in which a penalty is
affixed to a crime; or by the custom of the community and the
decisions of judges a common law gradually grows up in which
a penalty is so affixed. We must, I think, distinguish this stage,
that of the affixing of the penalty, from that of its infliction, and
we may ask on what principles the state or its officials should act
at each stage.
At the earlier stage a large place must be left for considera-
tions of expediency. We do not claim that laws should be
made against al! moral offences, or even against all offences by
men against their neighbours. Legislators should consider
such questions as whether a given law would be enforced if it
were made, and whether a certain type of offence is important
enough to make it worth while to set the elaborate machinery
of the law at work against it, or is better left to be punished
by the injured person or by public opinion. But even at this
stage there is one respect in which the notion of justice,
as something quite distinct from expediency, plays a part
in our thoughts about the matter. We feel sure that if a
law is framed against a certain type of offence the punish-
ment should be proportional to the offence. However
strong the temptation to commit a certain type of offence
may be, and however severe the punishment would therefore
have to be in order to be a successful deterrent, we feel certain
that it is unjust that very severe penalties should be affixed to
very slight offences. It is difficult, no doubt, to define the
nature of the relation which the punishment should bear to the
crime. We do not see any direct moral relation to exist between
wrong-doing and suffering so that we may say directly, such
and such an offence deserves so much suffering, neither more
nor less. But we do think that the injury to be inflicted on the
offender should be not much greater than that which he has
inflicted on another. Ideally, from this point of view, it should
be no greater. For he has lost his prima facie rights to life,
liberty, or property, only in so far as these rested on an explicit
or implicit undertaking to respect the corresponding rights in
others, and in so far as he has failed to respect those rights.
But laws must be stated in general terms, to cover a variety
of cases, and they cannot in advance affix punishments which
shall never be greater than the injury inflicted by the wrong-
doer. We are therefore content with an approximation to what
is precisely just. At the same time we recognize that this, while
it is a prima facie duty, is not the only prima facie duty of the
legislator; and that, as in the selection of offences to be legis-
lated against, so in the fixing of the penalty, he must consider
expediency, and may make the penalty more or less severe as it
dictates. His action should, in fact, be guided by regard to the
prima facie duty of injuring wrong-doers only to the extent that
they have injured others, and also to the prima facie duty of
promoting the general interest. And I think that we quite
clearly recognize these as distinct and specifically different
elements in the moral situation. To say this is not to adopt a
compromise between the intuitionist and the utilitarian view;
for it can fairly be claimed that one of the duties we apprehend
intuitively is that of promoting the general interest so far as
we can,
When the law has been promulgated and an offence against
it committed, a new set of considerations emerges. The ad-
ministrator of the law has not to consider what is the just
punishment for the offence, nor what is the expedient punish-
ment, except when the law has allowed a scale of penalties
within which he can choose. When that is the case, he has still
to have regard to the same considerations as arose at the earlier
stage. But that, when the penalty fixed by law is determinate,
this and no other should be inflicted, and that, when a scale of
penalties is allowed, no penalty above or below the scale should
be inflicted, depends on a prima facie duty that did not come in
at the earlier stage, viz, that of fidelity to promise. Directly, the
law is not a promise: it is a threat to the guilty, and a threat is
not a promise. The one is an undertaking to do or give to the
promisee someuung mutually understood to be advantageous
to him; the other, an announcement of intention to do to him
something mutually understood to be disadvantageous to him,
Punishment is sometimes justified on the ground that to fail to
punish is to break faith with the offender. It is said that he has a
right to be punished, and that not to punish him is not to treat
him with due respect as a moral agent responsible for his actions,
but as if he could not have helped doing them. This is, however,
not a point of view likely to be adopted by a criminal who
escapes punishment, and seems to be a somewhat artificial way
of looking at the matter, and to ignore the difference between
a threat and a promise,
But while the law is not a promise to the criminal, it is a
promise to the injured person and his friends, and to society.
It promises to the former, in certain cases, compensation, and
always the satisfaction of knowing that the offender has not
gone scot-free, and it promises to the latter this satisfaction
and the degree of protection against further offences which
punishment gives. At the same time the whole system of law
is a promise to the members of the community that if they do
not commit any of the prohibited acts they will not be punished.
Thus to our sense that prima. facie the state has a right to
punish the guilty, over and above the right which it has, in the
last resort, of inflicting injury on any of its members when the
public interest sufficiently demands it, there is added the sense
that promises should prima fade be kept; and it is the combina-
tion of these considerations that accounts for the moral satis-
faction that is felt by the community when the guilty are pun-
ished, and the moral indignation that is felt when the guilty
are not punished, and still more when the innocent are. There
may be cases in which the prima fade duty of punishing the
guilty, and even that of not punishing the innocent, may have
to give way to that of promoting the public interest. But these
are not cases of a wider expediency overriding a narrower, but
of one prima fade duty being more obligatory than two others
different in kind from it and from one another.

A STUDY of the meaning of 'good' and of the nature of

goodness should begin by recognizing that there is a wide
diversity of senses in which the word is used. The first distinc-
tion, perhaps, to be drawn is that between (A) the adjunctive
or attributive use of the word, as when we speak of a good
runner or of a good poem, and (JB) the predicative use of it, as
when it is said that knowledge is good or that pleasure is good.
It is evident that in ordinary usage the first meaning—that of
'good of its kind'—is much the commoner; it appears also to
be the earlier.1 Within the attributive use of the word we may
distinguish (i) its application to persons, and (2) its application
to things. In case (i) the root idea expressed by "good* seems
to be that of success or efficiency. We ascribe to some one a
certain endeavour, and describe him as a good so-and-so if we
think him comparatively successful in this endeavour. It might
be thought that in certain cases (e. g. *a good singer*, "a good
doctor') another idea is in our minds, viz. that the person in
question ministers to our pleasure, or to our health—in general
to the satisfaction of some desire of ours. But our pleasure or
our health comes in only incidentally in such cases; it comes in
just because the endeavour we are imputing to the person in
question is the endeavour to give us pleasure or to improve
our health. It does not, therefore, it would appear, form part of
the general connotation of 'good* when thus used. We can in
this same sense call a man *a good liar', not because he con-
tributes to the satisfaction of any of our desires, but because we
think him successful in what he sets out to do.
In case (2) there appear to be various elements included in
what we mean by 'good'. We seem to mean in the first place
(a) 'ministering to some particular human interest'. A good
knife is essentially one that can be successfully used for cutting,
a good poem one that arouses aesthetic pleasure in us. But
there is also here (b) the notion that the thing in question is one
« Ar.£jP.,s.v,'Good*.
in which the maker of it has successfully achieved his purpose
—a notion which might be called the 'passive' counterpart of
the notion explained under (i). As a rule both the notions (a)
and (£) appear to be involved in our application of 'good' to
anything other than persons; but sometimes the one and some-
times the other predominates. There is, however, (c) a third
element, less seriously intended, in our application of 'good*
to non-persons, When we speak of a good lie or of a good
sunset we are half-personalizing lies and sunsets and thinking
of this particular lie or sunset as succeeding in that which all
lies or sunsets are trying to achieve; i. e. we are, not quite
seriously, transferring to non-persons the meaning of 'good*
appropriate to persons.
Further, we have to note that 'good' in its application to
persons has a special sense in which it stands for moral excel-
lence. This is the case whether we emphasize the adjective or
the noun in the phrase 'a good man*. Both *a good man', as
opposed to a strong, clever, handsome, &c., man, and 'a good
man', as opposed to a good poet, plumber, scholar, Sec., stand
for moral excellence. The tendency to limit 'good* to the
meaning 'morally good* seems not to be involved in the original
connotation of the word, which is originally expressive of
indefinite commendation.1 The limitation seems to me to have
arisen in the following way. Mankind has, in an unsystematic
way, reflected a good deal on the question, what things are
good in themselves, intrinsically good, and has corne to think
that certain dispositions (of which the most conspicuous are
conscientiousness and benevolence) are the things that are
most certainly and in the highest degree good in themselves;
and it has tended more and more to adapt the adjunctive use of
the word to the predicative use, and apply it par excellence to
men characterized by such dispositions.
It is to be noted that 'good* in the sense of'good of its kind*
is doubly relative. It is in the first place relative to the kind-—
to what the kind is aiming at (when the word is applied to
Cf, the tf.E.D.'* primary definition of the worth "The most general adjective of
commendation, implying the existence in a high, or at least satisfactory, degree of
characteristic qualities which are either admirable in themselves or useful for some
persons) or to the activity which produces the kind or to the
interest which the kind subserves (when the word is applied to
things). When we call a person or a thing *a good so-and-so'
we do not imply that it is necessarily good in any respect other
than that expressed in the noun. What is a good x may be a bad
y or f. And 'good of its kind' is relative in a further sense, viz.
that it is comparative. We have in mind what we suppose to be
a rough average of the excellence of the members of the kind,
and we call anything better than this good and anything worse
than it bad, not implying that there is any fixed neutral point
at which what is good ends and what is bad begins. 'Good* in
this usage means 'better than the average* or perhaps 'consider-
ably better than the average", and 'bad* 'worse than the average'
or 'considerably worse than the average'. Whether 'good* is
used with reference to successful endeavour or to utility, we do
not imply, nor usually suppose, that there is a definite line
between success and failure, or between utility and inuttlity,
and that the things we call good are in any other than a com-
parative sense good, or those we call bad in any other than a
comparative sense bad.
Finally we may note that "good* in the sense of'good of its
kind* may be applied not only to an individual which is a good
instance of its species, but also to a species which is a good
species of its genus. We may say not only 'that is a good son-
net', but 'the sonnet is a good poetical form*.
We may turn now to the predicative use of 'good'. But it
must be noted that the grammatical difference is not a sure clue
to the difference of usage. Often when we say 'x is good* we
mean that it is a good so-and-so, and the universe of discourse
makes it clear what noun is to be understood. What I wish to
call attention to now is the cases in which there is no such
implication, as when it is said that 'courage is good* or 'pleasure
is good*. In such a usage, "good* is not relative in either of the
senses just pointed out. We do not mean that courage or
pleasure is a successful or useful instance of a species, or species
of a genus, nor do we mean that it is merely comparatively
good, rising above the average of its kind. In botli these re-
spects "good* in this usage is an absolute term.
Within this usage, however, several varieties have to be
distinguished, (i) m the first place, 'good' may here still mean
'useful*. A hedonist may call virtue good, though he means
only that it is conducive to pleasure. But while in calling a
thing good of its kind (when this refers to usefulness) we mean
simply that it conduces to the end things of its kind are meant
to conduce to (so that we may call a particular poison gas a
good poison gas, whatever we may think of the value of the
ends subserved by poison gases), in calling a thing good ab-
solutely (though still in the sense that it is useful) we mean
that it is a means to an end which is good; i.e. 'good* in this
usage is a complex notion implying both a causal relation
between the thing judged good and a certain effect, and the
goodness of the effect. Thus this usage points directly to
another (2), viz. that in which 'good' means 'intrinsically good',
I use this phrase rather than 'good as an end*, because the latter
phrase taken strictly would imply that the things referred to
are good only when desired, and therefore only when non-
existent. But that which is intrinsically good is not good only
when it is desired. If it is a thing good to be desired when not
yet existent, it is also a thing to be approved when it exists and
to be regretted when it has perished, and its goodness is no
more closely connected with the first of these attitudes than
with the other two. The intrinsically good is best defined as
that which is good apart from any of the results it produces.
(3) There is a sense of 'good* which Professor Moore has
distinguished from that conveyed by the expression 'intrinsi-
cally good*, as a narrower sense than this. He points out 1 that
hedonistic utilitarianism 'does not assert that pleasure is the
only thing intrinsically good, and pain the only thing Intrinsi-
cally evil. On the contrary, it asserts that any whole which
contains an excess of pleasure over pain is intrinsically good,
no matter how much else it may contain besides; and similarly
that any whole which contains an excess of pain over pleasure
is intrinsically bad.' What that theory asserts is that pleasure is
the only thing 'ultimately good* or 'good for its own sake*.
Both 'intrinsically good* and 'ultimately good* imply that the
* Etluct, 73.
thing in question would be good, even if it existed quite alone.
"We may, in short, divide intrinsically good things into two
classes: namely (i) those which, while as wholes they are in-
trinsically good, nevertheless contain some parts which are not
intrinsically good; and (2) those which either have no parts at
all, or, if they have any, have none but what are themselves
intrinsically good.* And he uses 'ultimately good* to denote
the second of these classes, 'Good throughout* would express
more obviously the same meaning.
The distinction is an important one. A whole, for instance,
which contains good elements and indifferent ones, but none
that are bad, is good 'apart from its consequences* and 'would
be good even if it existed quite alone", and is thus intrinsically
good, in the sense defined. But if the indifferent elements not
only are themselves indifferent but do not contribute to the
goodness of the whole, the whole is good not for its own sake
but for the sake of its good elements, and is thus not ultimately
good, in the sense defined. It might seem as if the distinction
answered to Aristotle's distinction between that which has a
certain attribute strictly gua itself (§ aw«J), and that which has
it in virtue of an element of itself (KOTO, pesos'); while the
wider distinction between the intrinsically good and the good
as a means answers roughly to Aristotle's wider distinction
between that which has an attribute in virtue of itself and that
which has it in virtue of a concomitant (*ca$' aura and Kara.
avfifteftrjKos). But Professor Moore would not accept this
identification, for his well-known doctrine of organic unities
states that elements in themselves indifferent or bad may yet
contribute to the goodness of a whole in which they occur.
Thus a whole containing, say, one good and one indifferent
element may have a goodness greater than that of its good
element, and is then not simply good 'in virtue of a part of
itself, while nevertheless it is not 'ultimately good*.
The importance of this doctrine is difficult to assess; for it
must be admitted that clear instances of 'organic unities* in this
sense are rather hard to discover. At first sight, one would say
that the clearest examples are to be found in the regions of
aesthetic and of economic value, (a) In the aesthetic region,
it is a familiar fact that some detail of a poem, say, or of a pic-
ture, which if it stood alone would have little or no aesthetic
value, yet contributes greatly to the effectiveness of the whole.
But if the view to be suggested later1 is true, that beauty is not
a form of intrinsic value, but only the power in an object of
evoking something that has value, the aesthetic experience,
then a beautiful object is not a case in point since the only value
that it or any of its parts has is an instrumental value. (t>) In the
economic region, it is a familiar fact that a pair of boots is
worth more than twice as much as a single boot, and an as-
sembled machine much more than the parts when unassembled.
But here the values in question are still more obviously instru-
mental, not intrinsic, and therefore not an illustration of the
Professor Moore's examples are not very convincing. Take
for instance his first illustration.2 *It seems to be true that to be
conscious of a beautiful object is a thing of great intrinsic value;
whereas the same object, if no one be conscious of it, has cer-
tainly comparatively little value, and is commonly held to have
none at all. But the consciousness of a beautiful object is
certainly a whole of some sort in which we can distinguish as
parts the object on the one hand and the being conscious on
the other. Now this latter factor occurs as part of a different
whole, whenever we are conscious of anything; and it would
seem that some of these wholes have at all events very little
value, and may even be indifferent or positively bad. Yet we
cannot always attribute the slightness of their value to any
positive demerit in the object which differentiates them from
the consciousness of beauty; the object itself may approach as
near as possible to absolute neutrality. Since, therefore, mere
consciousness does not always confer great value upon the
whole of which it forms a part, even though its object may have
no great demerit, we cannot attribute the great superiority of
the consciousness of a beautiful thing over the beautiful thing
itself to the mere addition of the value of consciousness to that
of the beautiful thing. Whatever the intrinsic merit of con-
sciousness may be, it does not give to the whole of which it
pp. 117—31. * Princtpia Ethica, 18-9.
forms a part a value proportioned to the sum of its value and
that of its object.' This analysis can surely not be accepted.
Consciousness, by which I think Professor Moore means
apprehension, is a state of a mind, and does not include its
object (say, a body) as a part of itself. The only whole which
could be said to include consciousness plus its object is the
whole (if it can be called a whole) consisting of the object plus
the consciousness of it. The true analysis of the consciousness
of a beautiful object, it would seem, is not into consciousness
plus the beautiful object, but into (a) its being an instance of
consciousness in general, and (£) its being an instance of con-
sciousness of something beautiful. And it seems to owe its
whole value to the second of the facts named.
Or again take another of Professor Moore's examples.1 'If
we compare the value of a certain amount of pleasure, existing
absolutely fy itself^ with the value of certain "enjoyments",
containing an equal amount of pleasure, it may become apparent
that the "enjoyment" is much better than the pleasure, and
also, in some cases, much worse. In such a case it is plain that
the "enjoyment" does not owe its value solely to the pleasure
it contains, although it might easily have appeared to do so,
when we only considered the other constituents of the enjoy-
ment, and seemed to see that, without the pleasure, they would
have had no value. It is now apparent, on the contrary, that the
whole "enjoyment" owes its value quite equally to the presence
of the other constituents, even though it may be true that the
pleasure is the only constituent having any value by itself.*
The situation here is that there are pleasurable states of mind
(e. g. enjoyments of beauty) which are judged to have more
intrinsic value than equally pleasurable states which were merely
pleasurable states could have. And Professor Moore seems
willing to admit that the element other than pleasantness in the
first kind of states of mind may have no value, or a value less
than the excess value of these states over the merely pleasurable
ones. If this be so, the case would certainly illustrate the doc-
trine of organic unities. But it seems at least arguable that the
element, other than pleasure, in the complex state—the element
Prindfia Ethica, iSi,
of insight, or whatever we may prefer to call it—has great
intrinsic value, enough to account entirely for the superior
value of the whole in which it is an element. And if so, the case
will not illustrate the doctrine of organic unities.
There seems, however, to be at least one case which illus-
trates the doctrine. Few people would hesitate to say that a
state of affairs in which A is good and happy and B bad and
unhappy is better than one in which A is good and unhappy
and B bad and happy, even \{A is equally good in both cases,
B equally bad in both cases, A precisely as happy in the first
case as B is in the second, and B precisely as unhappy in the
first case as A is in the second. The surplus value of the first
whole arises not from the value of its elements but from the
co-presence of goodness and happiness in one single person,
and of badness and unhappiness in another. And it is probable
that the principle has other applications, though it is hard to be
sure of these in detail.
The importance of the doctrine so far as its application goes
is somewhat doubtful. But its truth in the abstract seems un-
questionable. We have no right to assume that the value of
a whole is precisely equal, to the sum of the values of its elements
taken separately. It may owe some of its value to the co-
presence of certain of its elements in certain relations to one
another; and this co-presence of its elements cannot fairly be
called another element and thus taken to justify us in saying
that the value of the whole is the sum of the values of its ele-
In view of his doctrine of organic unities, Professor Moore
holds that there is yet another sense of good, (4), that must be
recognized. 'When we say that a thing is "good" we may
mean either (i) that it is intrinsically good or (a,)1 that it adds to
the value of many intrinsically good wholes or (3) that it is
useful or has good effects.'z The second of these three mean-
ings is properly distinguished both from the meaning 'intrinsi-
cally good* and from the meaning 'instrumentally good', and
might be called 'contributively good*. But it may be doubted
if it is a sense in which the word 'good' is often actually used.
= Our (4). * Ethics, 150.
For when a whole A-\-B is thought to be intrinsically good,
its goodness is usually taken to be either equal to that of one of
its elements A, B not being good in any sense (except perhaps
the instrumental sense), or if the goodness of the whole is taken
to be greater than that of A, an intrinsic goodness accounting
for the excess goodness of the whole is ascribed to £t. In fact,
in so far as the clear recognition of the principle of organic
unities is novel, this fourth sense of 'good* cannot have been
any one of the ordinary meanings of good, though many of the
things that have been called good may in fact have been good
only in this sense.
It is, I think, clear that it is the predicative rather than the
attributive senses of 'good* that are most important for philo-
sophy. And of the predicative senses, the first or instrumental is
clearly a complex notion including (a) the notion of a causal rela-
tion between something and something else, and ($) the notion
of the intrinsic goodness of the effect. It contains nothing but
these two elements. And of these two, it would be foreign to
our purpose to embark upon a discussion of causality. We are
left therefore with the notions of the intrinsically good, the
ultimately good, and the contributively good. Further, feeling
uncertain about the application in fact (though I do not doubt
the truth) of the principle of organic unities, I regard the things
that are intrinsically but not ultimately good as owing, gene-
rally speaking, their value to those elements in them that are
ultimately good; as being good, in fact, 'in virtue of a part of
themselves', the other parts of them being irrelevant to their
goodness. In ethics we have to take account of wholes that are
intrinsically but not ultimately good. For it is certain that when
we act we produce, along with any intrinsically good or bad
results that we produce, many others that are neither (e. g.
states of bodies). What we have to choose between the pro-
duction of is not states of affairs all of whose elements are
intrinsically good or intrinsically bad, but states of affairs many
of whose elements are neither. But these elements will (except
in so for as they may be contributively good) afford us no reason
for choosing to produce one such whole state rather than
another. And similarly in the parts that are intrinsically but
not ultimately good, the sub-parts that are not intrinsically
good will afford us no reason for trying to produce such parts,
What we must concentrate our attention upon is the elements
that are ultimately good; the others must be treated as mere
inevitable accompaniments of these. The notion of the ulti-
mately good—the notion, that is to say, of that which is good
strictly for its own sake and neither for the sake of its results
nor for the sake of an element in itself—is thus the central and
fundamental one,
But whatever is ultimately good is also intrinsically good,
i. e. is good apart from its consequences, or would be good even
if it were quite alone. We must make sure therefore that we
understand (as well as we can) the nature of intrinsic goodness.
And if we can once understand this, there will be little to add
about the further feature which distinguishes the ultimately
from the merely intrinsically good. We have simply to add
that the ultimately good as distinct from the merely intrinsically
good contains no element that is not intrinsically good.

E is round the question of the intrinsically good that the

hief controversies about the nature of goodness or of value
revolve. For most theories of value may be divided into those
which treat it as a quality and those which treat it as a relation
between that which has value and something else—which is
usually but not always said to be some state of a mind, such as
that of being pleased by the object or desiring it or approving
of it or finding its desire satisfied by it. And it seems clear that
any view which treats goodness as a relation between that
which is good and something else denies that anything is in-
trinsically good, since by calling a thing intrinsically good we
mean that it would be good even if nothing else existed. One
of the advocates of a relational view of value, Professor Perry,
seeks to maintain that a relational view does not involve the
dental of intrinsic value, which he evidently thinks would be a
consequence hostile if not fatal to his view. 'A . . . serious
objection' to his theory, he says,1 'is based upon the nature
of intrinsic value. We judge a thing to be intrinsically good
"where we judge, concerning a particular state of things,
that it would be worth while—would be *a good thing'—that
that state of things should exist, even if nothing else were to exist
besides, either at the same time or afterwards"/2 If a thing
derives value from its relation to an interest taken in it, it would
seem impossible that anything whatsoever should possess
value in itself. But in that case value would seem always to be
borrowed, and never owned; value would shine by a reflected
glory having no original source.
'The question', he continues, 'turns upon the fact that any
predicate may be judged synthetically or analytically. Suppose
that "good" were to be regarded as a simple quality like yellow.
It would then be possible to judge either synthetically, that the
primrose was fair or yellowj or, analytically, that the fair,
1 z
A General Theory ofFalut, 131, G. £. Moore, Edict, 161,
yellow primrose was fair or yellow. Only the fair, yellow
primrose would be fair and yellow "even if nothing were to
exist besides". But the logic of the situation is not in the least
altered if a relational predicate is substituted for a simple quality;
indeed it is quite possible to regard a quality as a monadic (a
single term) relation. Tangential, for example, is a relational
predicate; since a line is a tangent only by virtue of the peculiar
relation of single-point contact with another line or surface.
Let R' represent this peculiar relation, and At S, two lines,
One can then judge either synthetically, that (A) R'(B); or,
analytic-ally, that (A) R (J?) is R1. Similarly, let S represent
an interested subject, 0 an object, and R* the peculiar relation
of interest taken and received. We can then judge either syn-
thetically, that (O) R* (S); or, analytically, that (0) R* (S) is
J?'. In other words, one can say either that 0 is desired by 5",
or that O-desired-by-J" is a case of the general character "de-
I assume, as it seems necessary to assume in order to make
the example relevant, that 'fair* here = "beautiful*, and that
beauty is taken as a species of goodness. Professor Perry is
evidently taking 'yellow' to be a simple, non-relational quality,
and holding 'good* to be a relational one, viz. = 'object of
interest to some one' (loosely represented by 'desired-by-S'),
I am in doubt about the meaning of 'only' in the sentence
'Only the fair, yellow primrose would be fair and yellow "even
if nothing were to exist besides".* (i) 'Only* may mean *yet*.
If so, Professor Perry is admitting that the fair yellow primrose
would be yellow, and would be beautiful if beauty were a non-
relational quality, even if there were nothing else in the world.
And if this be so, that constitutes a vital difference between
such attributes, which would attach to their subject even if there
were nothing else in the world, and attributes such as 'desired-
by-S*, which certainly would not attach to 0 unless S existed
as well, (i) More probably, I think, 'only* means 'alone*; L e.
Professor Perry is saying that in contrast with the fair yellow
primrose, the primrose would not be fair and yellow if nothing
else were to exist, just as 0 would not be desired by S if S did
not exist as well as 0.
Now it is true that the primrose-could not be fair or yellow
if nothing but it existed. It could not be fair if its fairness did
not exist, nor yellow if its yellowness did not. But it is equally
true that the fair yellow primrose (which Professor >Perry con-
trasts with the primrose in this respect) could not be fair or
yellow if its fairness or its yellowness did not exist; and its
fairness and its yellowness are quite as different from the fair
yellow primrose as they are from the primrose, so that there is
no difference between the primrose and the fair yellow primrose
in this respect. But if yellowness, or fairness, is a non-relational
quality of the primrose, the primrose might be yellow or fair
though nothing but the primrose and its attribute of being
yellow or fair existed. On the other hand, if goodness is a
relational quality (say=object of interest to some one), nothing
could be good unless, besides it and its attribute of "being an
object of interest to some one', something else existed, viz, a
person to whom it is an object of interest. The essential differ-
ence would remain, that non-relational attributes can be pos-
sessed by subjects though nothing but the subjects and the
attributes exist; while relational attributes can be possessed by
subjects only if something besides both the subjects and the
attributes exists, viz, the things that form the other terms of
the relations. Thus if the definition of an intrinsic attribute as
one which its subject would possess if nothing other than the
subject existed, be amended into the form 'an intrinsic attribute
is one which the subject would possess even if nothing but the
subject and the attribute existed', it is evident that non-rela-
tional attributes are intrinsic and that relational attributes
cannot be so. If 'good*, then, be defined as Professor Perry
defines it, nothing can be intrinsically good. And his attempt
to get over the difficulty of the apparent necessity (for a rela-
tional view of value) of denying that anything has intrinsic
value, by means of the distinction between analytic and syn-
thetic judgements, comes to nothing. 'O-desired-by-5* is not
a different object which can truly be said to possess intrinsic
value when it is denied that any O apart from being desired
has intrinsic value. 'O-desired-by-J is good* is simply another
way of saying 'any 0 has value not in itself but by virtue of the
co-existence with it, and in a certain relation to it, of S', And
to say this is to deny intrinsic value to anything. And similarly
any other view which identifies goodness with or makes it
depend upon a relation between that which is good and some-
thing else, denies the existence of intrinsic value,
The theories which identify goodness with some relation
are bound to think of this either (I) as a relation between that
which is good and some or all of its elements, or (II) as a rela-
tion between some or all of its elements, or (III) as a relation
between it or some or all of its elements and something else.
Out of the many theories about the nature of goodness, I am
unable to think of any which belongs to type (I), and this type
need not, perhaps, be examined. There have been theories of
type (II), viz. those that identify the good with the harmonious
or coherent. With reference to any such view, the question
must first be asked whether it is meant (a) that goodness just is
coherence, or ($) that what is good is good because it is coherent.
Only the first of these views is strictly relevant here, where we
are inquiring what goodness is. The second view does not
answer this question; it leaves still open the question what is
the nature of the attribute goodness which coherent things
are said to have because they are coherent. Now the first view
seems to be clearly false. It is surely clear that, however close
a connexion there may be between coherence and goodness,
we never meaa% when we call a thing good, that it is coherent.
If this were what we meant, 'the coherent, and only the co-
herent, is good' would be a mere tautology, since it would be
equivalent to 'the coherent, and only the coherent, is coherent*;
but it is evidently not a mere tautology, but a proposition which
if true is very important The theory then, if it is to have any
plausibility, must be understood in the second form; and in
this form it is no answer to the question we are asking, what is
It may be well, however, to offer some comments on the
theory in its second form, even if this is not strictly relevant to
the present stage of our inquiry. In the first place it may be re-
marked that any such theory seems to start with the presump-
tion that there is some single attribute, other than goodness,
that makes all good things good, and that the only question is
what this attribute is. Now I agree that goodness is a conse-
quential attribute; that anything that is good must be good
either by virtue of its whole nature apart from its goodness, or
by virtue of something in its nature other than goodness. This
seems to me a very important fact about goodness, and one
that marks it off from most other attributes,1 But I cannot
agree that the presumption is that there is any one characteristic
by virtue of which all the things that are good, are good. If
conscientiousness and benevolence, for instance, are both good,
it is just as likely, initially, that conscientiousness is good
because it is conscientiousness, and benevolence good because
it is benevolence. Still, this must not be assumed to be the case,
any more than the opposite view must be assumed. We must
be prepared to consider on its merits any suggested general
ground of goodness. But when I ask myself whether conscien-
tiousness, or benevolence, for instance, can be held to be good
by virtue of the coherence of its elements, I have to ask what
the supposed elements are, and in what respect they are sup-
posed to cohere, and to these questions I find no clear answer
given by those who hold the theory. It would be more plausible
(though not, I think, true) to say that the goodness of con-
scientious or benevolent action depends on its coherence with
something outside it, e. g, with the whole system of purposes of
the agent, or of the society he lives in. But such a theory would
belong not to type (II), which we are examining, but to type
(III). Or again, suppose that one judges a particular pleasure
to be good, is it not clear that even if most and possibly all
pleasures are complex, it is not on account of its being a com-
plex united by the relation of coherence, but on account of its
having the felt character of pleasantness, that it is judged to be
good ?
When we turn to type (III), we find that the relation which
is identified with goodness (or else held to be what makes good
things good) is sometimes held to be necessarily a relation to a
mind, while sometimes this limitation is not imposed. I take
as a typical view of the latter kind one of which I owe my
» Cf. pp. 1*1-2.
knowledge to an article by Professor Urban.1 Professor Shel-
don, as reported by Professor Urban, holds that value is 'fulfil-
ment of any tendency whatever*. The essential objection to
this theory seems to me to be this. Empty "tendency* of any
reference to the aims of conscious beings (which it is the special
point of this theory to do), and what meaning is left for 'fulfil-
ment of any tendency"? What is left is the notion of a thing's
being under the influence of a certain force, and of its actually
passing into the state it would pass into if acted on by that force
alone. And who will say that this purely physical circumstance
is either identical or even coextensive with value ?
To this Professor Sheldon (as reported by Professor Urban)
answers: 'Good is no doubt a different notion from fulfilment,
and therefore appears to contain something not authorized in
the content of the latter notion. But that is because good or
value Ss the relation between the fulfilment (or furthering) and
the tendency, a relation uniquely and sufficiently determined
by the two.*2 To this it seems to me enough to reply that this
relation can exist, as much as anywhere else, in the case of
bodies acted on by physical forces, where no one would drearn
of applying the notion of good or value. If we must have a
relational theory of value, there seems to be much more plausi-
bility in the 'psychological' than in the 'ontoiogical* form of
the theory.
The 'psychological* theories as a rule take the form of hold-
ing that a thing's being good means either (A) that some person
or persons have some feeling towards it, or (j?) that some per-
son or persons think it to be good; and such views, or rather
those of the first type, have some initial attractiveness. (A) Our
judgements that certain things are good are in fact constantly
accompanied by feelings towards them'—feelings of pleasure,
and of regret for their absence; and this fact Is apt to lead to one
or other of two views, or more often perhaps to a mixture of
the two. One view is that by being objects of some such feeling
(let us say, adopting Professor Perry's comprehensive phrase,
by being 'objects of interest') things acquire a further character,
that of value. The other is that to have value is just to be an
' Journal of Philosophy, &c>, 1916,454. * Ib.
object of interest, and nothing more. I am rather in doubt how
to classify the view put forward in A General Theory of Value
by Professor Perry himself. Passages could be found in his book
to support the interpretation of him as holding the first viewj
e. g. those in which value is described as dependent on interest.
But on the whole it seems pretty clear that it is the second view
he wishes to maintain, 'The view*, he says,1 'may otherwise be
formulated in the equation: x is valuable == interest is taken in
x'; and immediately after, "Value is thus a specific relation into
which things possessing any ontological status whatsoever,
whether real or imaginary, may enter with interested subjects*
—i. e. the relation of being objects of interest to them.. Again,2
'Thus the question* (the question to which he provides an
answer) 'is the question, In what consists' (not, On what de-
pends) 'value in the generic sense ?*
If the fast interpretation be the true one, there remain diffi-
cult questions to which he provides no answer. If value is
something not consisting in, but depending on, being an object
of interest, what is value itself, and what is the nature of the
relation vaguely described as dependence? Is the relation a
causal one, or a logical one, and if neither of these, what is it?
To these no answer is suggested. But these questions need not
be pressed, for I fancy that Professor Perry would accept the
second interpretation as the true account of his view,
On this second interpretation, the theory is that 'good' and
"object of interest' are just different ways of expressing exactly
the same notion. But it is surely clear that this is not true. It is
surely clear that when we call something good we are thinking
of it as possessing in itself a certain attribute and are not think-
ing of it as necessarily having an interest taken in it. If when
we attend to something we are impelled to describe it as good,
it is surely not impossible to think that, though of course we
can only discover its goodness by attending to it, it had its
goodness before we attended to it and would have had it if we
had not attended to it. And again it is evidently possible to
think that some of the things in which an interest has been
taken have nevertheless been bad. But if 'good* and 'object of
A General Theory of Valm, 116. * Ik 118.
interest* meant exactly the same, it would be impossible to think
either of these two things which it clearly is possible to
think. The view, therefore, that 'good' and 'object of interest*
stand for the same notion must be given up. What the rela-
tional theory must maintain, if it is to be plausible, must be
something different; it must be that whereas most people think
that certain things have a characteristic, goodness, distinct from
that of being objects of interest, nothing has any such charac-
teristic. And then the question arises, what could have led
mankind to form this quite superfluous notion to which nothing
in reality corresponds ? • It is not as if the notion of goodness
were a complex notion formed, like such notions as that of
'centaur', by a play of fancy in which characteristics found
separate in reality are imagined to coexist; for there are no
characteristics of which 'good* can be said to be a compound,
We may, however, not merely ask how the notion could have
come into being if it were not the apprehension of a reality,"
We may claim that we are directly aware that conscientious
action, for example, has a value of its own, not identical with
or even dependent upon our or any one else's taking an interest
in it. Our reason informs us of this as surely as it informs us
of anything, and to distrust reason here is in principle to dis-
trust its power of ever knowing reality.
Another fatal objection to any theory which identifies good
with being an object of interest, or of any particular type of
feeling, becomes apparent when we ask by whom the interest
or the feeling is supposed to be felt. Some answers escape some
objections and others escape others, but each possible answer is
exposed to at least one fatal objection of its own. This ground
has been very fully covered by Professor Moore in an examina-
tion of the corresponding theories about 'right',2 and both in
the case of'right' and in the case of'good* his line of argument
seems to me unanswerable. Theories of this type are divisible
into those which identify goodness with the presence of some
feeling (i) in at least one person, no matter who he is, (2) in
the person who judges an object to be good, (3) in a majority
Cf, Cook Wilson's argument against the possibility of a fictitious 'simple idea*,
Statement andInference, u, 511—it. Ethics, chs, j and 4,
of persons of some class or other—say persons belonging to a
particular stage in the history of civilization, (4) in a majority
of mankind, or (j) in all mankind. To (i) there seem to be
four objections, (a) It surely can hardly be denied that, what-
ever feeling we select as the feeling involved—whether for
instance this be taken to be pleasure, or approval—a man may
doubt whether a certain thing is good, even when he does not
doubt that some one or other has had such a feeling towards it.
($) If what I mean when 1 call something good is that some one
or other has a certain feeling towards it, and if what any other
person means when he calls it bad is that some one or other has
an opposite feeling towards it, we should not be at variance,
because both propositions might be true. Yet if anything is
clear, it is that we do suppose ourselves to be making, incom-
patible statements about the object, (c) If something, without
changing its nature, at some moment aroused for the first time
the feeling in question in some mind, we should clearly judge
not that the object had then first become good, but that its
goodness had then first been apprehended. And (</) it might
be enough to ask whether any one finds it even possible to think
that goodness could be brought into being by the feeling of
some one or other, no matter how vicious or stupid or ignorant
he might be. It seems clear that by goodness we mean some-
thing at any rate more objective than that.
To the theory in form (2) the primary objection is identical
with objection (<J) above. If all I mean by saying that an
object is good were that it arouses a certain feeling in me, and
all you mean by saying that it is not good, or is bad, were that
it does not arouse that feeling, or arouses an opposite feeling,
in you, we should not be at variance, for we might both be
right. And objection (c) applies with just as much force to this
theory as to the previous one.
To the theory in form (3) it may be objected (a) that it will
follow that two people who claim to be representing the feel-
ings of majorities of different sets of persons will never be at
variance if they pronounce the same thing respectively good
and bad. Yet it is clear that even when two men belong to
different sets of persons, the feelings of a majority of which
they would on this view be claiming to represent, they believe
themselves to be making incompatible statements when they
call something respectively good and bad. Clearly therefore
what they claim to be expressing is not the feelings of different
majorities. But further (£) it is surely plain that there are cases
in which a man thinks something good, without thinking that
there is a majority of any class of men who have a certain feeling
towards it. Even if we think that a majority of persons at our
own stage of civilization, for instance, would have feelings like
ours if they attended to the object, we may feel sure that they
have not attended to it and therefore have not the feeling in
question towards it.
The theory in form (4) is not open to the fast objection made
to the previous theory. For any one who thought that a majority
of mankind had a certain feeling towards an object would be
at variance with any one who thought that they had not this
feeling, or had an opposite feeling. But objection (K) to theory
(3) applies with redoubled force to theory (4).
And finally, to theory (f) it applies with even greater force,
(jff) The second and remaining type of what I may call purely
subjective theories of good is that which holds that for me to
think an object good is to think that (i) some one or other, or
(2) I, or (3) a majority of some set of men, or (4) a majority of
mankind, or (5) all mankind, think it good. It is unnecessary
and would be tedious to examine these theories as fully as we
have examined those of type (A), It is enough to point out
that corresponding objections are equally fatal to them, and to
add a new objection fatal to all theories of type (B),
The objections to Ai apply equally to B i
» n » "^ « 5? „ .02
n n » -"3 J? » n **3
» » fi -"4 •>•> i> 55 -"4
» M » •" 5 n » » ** 5
But apart from these objections to special forms of theory (jff),
the whole theory has one absurdity common to all its forms.
It is perfectly evident that the meaning of 'Jf is good* cannot be
identical with the meaning of'some one (or I, or a majority of
some class of men, &c.) thinks that X is good', since it is identi-
cal with the meaning of only one element in the latter phrase,
Or, to put the same objection otherwise, to say that S thinks X
good leaves it an open question whether X is good. For
opinion has the characteristic, which feeling has not, of being
either true or false. If S thinks falsely that X is good, then X
is not good; and if S thinks truly that X is good, then X'$ being
good is neither identical with nor dependent on S's thinking
it good. In fact, while theory (A) deserves the most serious
consideration, and it is excessively hard to be sure whether one
is right in rejecting it or may not have been guilty of some
logical confusion, theory (B) may be rejected out of hand.
Professor Perry, as one might expect, repudiates it with vigour.

I turn to a reconsideration of theory (A) in the light of

Professor Perry's discussion. He divides all possible theories
of value into four types, according to the view they take of the
relation of value to interest, interest being identified 'with the
motive-affective life; that is to say, with instinct, desire, feeling;
these, and all their family of states, acts, and attitudes',1 'There
are four possible relations of value to interest. In the first place,
value may be, in its essential nature, quite irrelevant to interest,
... In the second place, value may be held to be the character
of an object which qualifies it to be an end; in other words,
that which implies, evokes or regulates interest... In the third
place, value may be assigned to the objects of certain duly
qualified interests, such as the final, harmonious, absolute, or
imperative interest. Finally, there is the simpler and more
comprehensive view, that value in the generic sense attaches
promiscuously to all objects of all interest.'2
I am not specially concerned with the two intermediate
views, and agree with many of Professor Perry's criticisms of
them. I am mainly interested in the first, which I believe to be
true,3 and in the fourth, which he believes to be true. He takes
as a typical expression of the first view Professor Moore's
* A General Theory of Value, 17. Ib.
> 1 think, of course, that a thing may arouse interest, and will arouse it in a well-
constituted mind, ttteute of its goodness. What I wish to deny is that its goodness
cither is or depends on its arousing interest.
remark 'my point is that "good" is a simple notion, just as
"yellow" Is a simple notion; that, just as you cannot, by any
manner of means, explain to any one who does not already
know it, what yellow is, so you cannot explain what good is*;
and he treats this view as being best 'understood as an extension
of that pan-objectivism which, having concluded that the so-
called "secondary qualities", such as colour, have as good a
title to extra-mental existence as the so-called primary qualities,
such as figure, sees no reason why the so-called "tertiary"
qualities, such as good, should not be assigned the same
There may be in some minds a connexion between a realistic
view of the secondary qualities and an objective view of good-
ness, but it should be pointed out that there is no necessary
connexion between the two views. For my own part, reflection
on the facts of perception and of its illusions forces me to think
that there is no such thing as objective colour, for example;
I arn driven to suppose that colour-sensation is a mental state
which is not perception of colour. But colour-sensation 2 is an
indubitable fact, and I can with a certain modification accept
Professor Moore's comparison, I can say that goodness is a
quality which can no more be defined in terms of anything
other than itself, than can the quality of the sensation which we
describe as being one of 'seeing yellow*. Whatever we may
think about the objectivity of colour, there can, I imagine, be
no doubt of the indefinability of the character of our sensation.
Thus the adoption of this comparison is in no way bound up
with an objective view of secondary qualities. Nor, again, do
I think of goodness as 'extra-mental*; for while I do not think
it is essentially for minds, I think it is essentially a quality of
states of mind.
Professor Perry's first criticism of the objective view of good
is that 'one who upholds this view of good must be prepared
to point to a distinct quale which appears in that region which
our value terms roughly indicate, and which is different from
A General Theory of Value, ^<),
* i. e. the experience which we habitually, whether rightly at (as I suggest) wrongly,
describe as that of seeing colour.
the object's shape and size, from the inter-relation of its' parts,
from its relation to other objects, or to a subject; and from all
the other factors which belong to the same context, but are
designated by words other than "good". The present writer,
for one, finds no such residuum.'1 The existence of such a
residuum is just the point at issue. So far we have only the word
of those who agree with Professor Moore that they do discern
in certain things a unique quality which can only be expressed
by the term 'goodness* or some synonym of 'goodness', and
the word of those who agree with Professor Perry that they do
not; and so long as the question is considered on these lines,
all that we can do is to invite others to contemplate, for instance,
conscientious action, and try for themselves whether they do
or do not discern such a quality in it. But Professor Perry is,
of course, not content with his ipse dbdt. He argues that if
goodness were an indefinable quality like yellowness its presence,
when it is present, should be equally self-evident; and he points
to the hesitancy of Professor Moore's report as to what things
are good, as showing that the presence of goodness is not equally
self-evident with that of yellowness. Here lie seems to be stress-
ing too much the analogy which Professor Moore has alleged to
exist between goodness and yellowness. The analogy exists
only in respect of the indefinability of both. It is not argued
that in other respects the two qualities are on all fours. In
particular, the one is apprehended (if apprehended at all) 2 by
sense-perception, the other by intelligence; and there is no
reason to anticipate that what is discerned by the intelligence
should be as easily discerned as what is discerned by sense-
perception. But Professor Perry exaggerates the difference
between the ease of discernment in the two cases. There is, he
says, 'no serious difference of opinion as to the distribution of
terms connoting empirical qualities. "Things wear them in
public, and any passer-by may note them." ' 3 But does not
yellow merge into green, and into orange, and are there not
* A General Theory of halite, 30.
This caution seems necessary in view of the doubt I have expressed on p. 86 as
to whether eofow is something apprehended at all. The sentence in the eext without
the parenthesis would state the defence which I believe Professor Moore would make
ofliisvicw, ' A General Thtary tf PW««, jo.
border-line in which it is extremely difficult to say whether
what we have before us is yellow or green, or whether it is
yellow or orange ? And if there are things about whose good-
ness there is room for difference of opinion, are there not other
things, such as conscientious action, whose goodness is matter
of general, agreement ?
But I should not like to rest the case for the indeinabilhy of
goodness merely on this argumentum ad hommem. It seems to
me more important to point out that the question whether the
presence of a given quality in some particular thing is easily
discerned has nothing to do with the question whether the
quality is indefinable. If two people differ, for instance, as to
whether a particular action is good, their differing implies, no
less than their agreeing would have done, that they mean by
"goodness* a definite quality; and their mere differing does not
imply that that quality is not indefinable any more than their
agreement would imply that it is. The questions of its de-
finability and of its discernibility are different and not logically
But if I attempt to vindicate Professor Moore's comparison
of goodness with yellowness as being like it an indefinable
quality, I do not wish (any more than I imagine he would) to
be thought to suppose that it is a quality in other respects like
yellowness. The most salient difference is that it is a quality
which anything that has it can have only in virtue of having
some other characteristic; as e.g. a conscientious act is good
in virtue of being conscientious. This I express later by de-
scribing it as a consequential and not a fundamental quality.1
Professor Perry turns next to mention Professor Laird's
presentment of the objective view. Professor Laird, he says,
'appeals to the fact that there is an immediate objectivity in the
appreciation of beauty, or in the admiration of conduct These
are not mere subjective states caiaedbj an object; they present
the object, clothed in its quality of charm or moral worth.*2
Professor Perry points out that there are many adjectives which
we apply to objects, and which therefore prima, facie might
appear to stand for qualities of objects apart from any relation
Cf, pp. 121—i. * A General Theory of lvalue, 31.
to persons, but which on examination turn out to refer simply
to the existence of some such relation,—ad jectives like 'coveted*,
'boresome*, 'tiresome*, 'hopeful*; and that on the other hand
adjectives like 'red* resist all attempts to localize them in the
subject and insist on being localized in the object. But it is
surely unfair to argue from words like 'coveted', 'boresome*,
'tiresome*, 'hopeful', which by their very formation point to a
relation between a subject and an object, and the word 'good',
which equally clearly points to nothing of the kind but to a
quality resident in the object itself, independent of any subject's
reaction to the object. As regards 'beautiful* I am, as I shall
point out later,1 inclined to agree that the fact that lies at the
back of our predications of it is simply the power something
has of producing a certain kind of emotion in us; and the fre-
quent use of such words as 'charming', 'delightful*, almost as
synonyms of 'beautiful* may be held to lend this view some
support. But it is surely a strange reversal of the natural order
of thought to say that our admiring an action either is, or is
what necessitates, its being good. We think of its goodness as
what we admire in it, and as something it would have even if
no one admired it, something that it has in itself. We could
suppose, for instance, an action of self-denial which neither the
doer nor any one else had ever admired. If now some one were
to become aware of it and admire it, he would surely pronounce
that it had been good even when no one had been admiring it,
Professor Perry makes the further objection that the 'objec-
tive* theory derives all its plausibility from its exponents* being
preoccupied with 'the aesthetic and contemplative values', and
that it precludes them from giving a comprehensive account
of all values. 'The most serious defect of this type of theory
is its failure to provide any systematic principle whatsoever.
There are as many indefinable values as there are feeling atti-
tudes, and since these are to be regarded as objective qualities
rather than as modes of feeling, there is nothing to unite them,
not even the principle of feeling. If "good" is a unique quality,
then so are "pleasant", "bad", and "ought". There is no way
of subsuming pleasant under good, or of defining the opposi-
' pp. izy-jo.
don of good and bad, or of subsuming both good and ought
under a more general category such as value. If, on the other
hand, value is defined in terms of interest, then the variability
of interest seems to account for both the unity and the diversity
of values.*1 His assumption, then, is that there must be some
single sense of 'valuable' in which the word is always used,
and his contention is that a subjective theory alone will serve
to assign such a single meaning and to show the relations
between the various specific kinds of value. And under the
heading of Valuable' he includes both things which would not
naturally be described as being valuable at all, and things which
we can surely recognize to have value only in fundamentally
different senses. Does any one really think that obligatoriness
is a special form of being valuable? 2 Is it not a liasty assump-
tion to assume that it is an instance of the same kind of thing
of which moral goodness or beauty is another instance? And
is it not clear that what we call economic values 3 are merely
instrumental values, different in kind from the goodness of
virtue or of pleasure? The assumption that there must be *a
general theory of value' applicable to value in all the senses of
that word seems to me to be unjustified.
At the same time, I am inclined to agree with Professor Perry
in one of his contentions, though not in what he (if I understand
him aright) seeks to deduce from it. He is seeking to find a
single thread of identity which unites all our applications of the
word good, and to infer from this that the word 'good* has the
single meaning which he assigns to it Now when I consider
the variety of meanings of 'good* indicated in the preceding
chapter—the predicative and the attributive use, the mean-
ings 'successful in his endeavour* and 'useful', the instrumental
and the intrinsic sense—though I cannot agree that what we
mean in all or any of these cases by "^ is good* is 'Xis an object
of interest to some one*, I am inclined to think that the only
thread that connects our application of the word in all these
senses—i. e. the only common fact that is present whenever we
use the term "good*—is that in each case thejuJger has some
feeling of approval or interest towards what he calls good. But
1 l
A General TAtory of Falut- 34. Ib. J Ib.
this in no way proves that we are always using 'good* in the
same sense. The senses 'intrinsically good* and 'useful' appear
to me entirely different, though whether we use the word good
in one or the other we have in both cases a feeling of approval
or interest towards what we call good. What common thread
there is, is one that connects not the various meanings of good,
but our use of it in these various meanings. The attempt to find
a common thread in our application of the term is not what I am
chiefly interested in. What I am interested in, and what I can-
not but think to be the more important question for philosophy,
is whether there is not a sense of good in which it can be applied
to things not as meaning that they are successful or useful
members of a class, and not as meaning that they are instru-
mental to a good beyond themselves, but as meaning that they
are good in themselves. And it is surely plain that when we
state, for instance, that courage is good, this is what we mean—
even if some one may maintain that we are mistaken in making
this statement. I have tried to do some justice, briefly, to
the other senses, in the preceding chapter, and from that
point onwards I have been interested solely in this other and
more fundamental sense of 'good*. And of this I feel pretty
clear, that though our applications of it are always accompanied
by an interest in what we thus call good, the existence of that
interest is not what we assert when we so describe things.
Professor Perry turns next to consider in detail Professor
Moore's argument for the indefinability of 'good'. He quotes
the remark "it would be absolutely meaningless to say that
oranges were yellow, unless yellow did in the end mean just
"yellow" and nothing else whatever-—unless it were absolutely
indefinable."' And to this, taken alone, his objection is well
founded. 'It is not meaningless', he points out, 'to say that
"the conception of substance is pre-historic", or that "the
painting is post-impressionistic", or that "the argument is
circular"; and yet in these cases the assigned predicates are de-
finable."2 The statement 'oranges are yellow* certainly is mean-
ingless unless 'yellow* has in this statement a single self-identical
meaning. In a sense 'yellow' must mean yellow and nothing
' G. E. Moore, Principles Etkka, 14. A General Theory of Value, 3,5.
else whatever. But this does not show it to be indefinable. For
if we could correctly define 'yellow* as, say, 'x which isj*, we
should not be saying that 'yellow' means anything other than
'yellow', for 'x which isy* would be just what yellow is. But,
it might be said in support of the view that 'good* is obviously
indefinable, there is a great difference between "good* and such
attributes as 'pre-historie', "post-impressionistic', 'circular*. If
a term is definable, i. e. stands for a certain complex,1 we can
use the term intelligently and intelligibly only if we have the
definition to some extent before our minds; and we have in fact
at least rough definitions of such terms in our minds when we
use them. On the other hand, the fact that we use the term
'good* intelligently and intelligibly without having any defini-
tion of it in our minds shows that it is indefinable.
Professor Moore uses an argument of somewhat the same
type when he argues, against any attempt to define 'good*,
that given any set of concepts not containing good, it is always
possible to inquire whether a thing answering to this set of
concepts is good,2 Suppose some one claims that 'being-desired-
by-anybody is being good*, this claim is met by the fact that
even if we know that war is desired by some people, we may
still doubt whether it is good.
Both of these arguments amount, 1 think, to saying that if
'good* stood for any complex (as on any relational theory it
does), we ought, if we use the word intelligently, to have in
our minds the notion of a definite relation between definite
things. It seems to me clear that we have no such notion in our
minds when we use the word in ordinary discourse.
But I cannot be sure that this entirely settles the question.
For there seem to be cases in which we seek for the definition
of a term and finally accept one as correct. The fact that we
accept some definition as correct shows that the term did some-
how stand for a complex of elements; yet the fact that we are
I should explain that I mean i>y a complex here a complex of elements co-ordinate
in respect of universality, in distinction front another class of terms which mlgkt be
called complex, and which are indefinable, viz. those that involve elements not co-
ordinate in respect of universality, as 'red* involves both colour and redness, Cf. Cook
Wilson, Statement and Inference, u. 502-4.
Pnnclpla £»thica, 15*
for some time in doubt whether the term is analysable, and if so,
what the correct analysis is, shows that this complex of elements
was not distinctly present to our mind before, or during, the
search for a definition. It appears as if we cannot avoid recog-
nizing that there is such a thing as using a term which implicitly
refers to a certain complex, while yet the complex is not ex-
plicitly present to our minds. And in principle this might, it
seems, be true of 'good'. The absence of an explicit reference
to a complex in our ordinary use of the term should therefore
not be taken as necessarily implying that the term is indefinable,
nor, in particular, as excluding the possibility of its standing
for a relation. The method should, I think, rather be that of
attending to any proposed definition that seems at all plausible.
If it is the correct definition, what should happen is that after
a certain amount of attention to it we should be able to say,
'yes, that is what I meant by "good" all along, though I was
not clearly conscious till now that it was what I meant*. If on
the other hand the result is that we feel clear that 'that was not
what I meant by good', the proposed definition must be re-
jected. If, after we have examined all the definitions that possess
any initial plausibility, we have found this negative result in
every case, we may feel fairly confident that 'good' is indefi-
nable. And there is no initial presumption that it is definable.
For it seems clear that there could be no complex entities unless
there were some simple ones; and, in a universe so various as
the universe is, there is no reason to suppose that the simple
entities are few in number.
In the process of criticizing proposed definitions of a term,
there, are two moments. Perhaps the most obvious ground for
rejection of a definition is that we are able to point to things of
which the term is predicable but the definition not, or vice versa.
And any one will be able without difficulty to think of defini-
tions of 'good' that have been proposed, which corne to grief
on one or other of these two objections. But even when the
denotations of the term and of the definition coincide (or when
we cannot be sure that they do not), we can often see that a
proposed definition does not express what we mean by the
term to be defined. It would be on this ground, for example,
that we should reject a definition of "equilateral triangle' as
'triangle with all its angles equal'. And it is on this ground
that most of the proposed definitions of 'good* can be rejected
—many of the metaphysical definitions, such as those which
identify goodness with comprehensiveness or with reality;
and many of the psychological definitions, such as those which
identify it with being productive of pleasure or with being an
object of desire. The point is not that the proposed definition
is not seen at first sight to be true, or that it needs inquiry, but
that it does not survive inquiry.
Professor Perry's own criticism of Professor Moore takes
the following form. Suppose that 'good* be defined as 'desired
by some one'. This definition is disproved, says Professor
Moore, by the fact that even if war is desired by some one, it is
still possible to inquire whether war is good. Professor Perry
seems to admit this as fatal to the proposed definition, for he
proposes to substitute for it what he evidently thinks of as a
different definition, 'good in some sense=desired by some one*.
And he endeavours to turn the edge of Professor Moore's
objection by saying that the correctness of this definition is
quite compatible with our still being able to inquire (as we
evidently can) whether war, if it is good in this sense, is also
good in some other sense, e, g. desired by all men, or obligatory,
or beautiful.1
This seems to me strangely to miss the point. No one would,
I suppose, dream of objecting to the equating of "good in some
sense' with "desired by some one* on the ground that war though
desired by some persons is not desired by every one, or not
beautiful, or not obligatory. The objections are (i) that, even
though desired by some persons, war is not in any sense good
(though there may be elements in it that are good), and (2)
that, even if it were In some sense good, what would be meant
by calling it good is most certainly not that it is desired by
some one.
Professor Perry further tries to base an argument for the
relativity of good to the interests of individuals, on the fact
'that the question may be submitted once again to each mdi~
> A Gtatral Theory of Vnlmt 36-7.
victual judge', "If when a given object a is already acknowledged
to be good the question of its goodness is nevertheless put to a
subject M, the question is assumed to refer to the special sense
of good which is relative to M.'l This would surely be a non-
sensical procedure. If the goodness of a is already "acknow-
ledged', i. e, admitted by both the persons involved, there is no
sense in the one asking the other whether the object is good;
and anything that the other may say such as 'I desire it' or 'I
don't desire it* has no relevance to the question whether it is
good, which has already ex hypothesi been settled in the
The advocates of the view he is criticizing are, Professor
Perry points out, anxious to secure for 'good' a meaning which
shall 'provide judgments of value with a common object which
will determine their truth or falsity*. And for this purpose, he
insists, "an interest is as good an object as any other. The fact
that M takes an interest in a, consists in a relation of a to M;
but this fact itself is not relative to M's judgment about it, or
to the judgment of any other subject.*2 There is, in fact, & great
difference between a view which makes the goodness of an
object depend on a subject's judgement that it is good, and one
which makes it depend on his interest in it. The former view
is one that will not stand a moment's examination; the latter is
one that does provide judgements of value with some reality
to judge about, and that therefore requires serious considera-
tion. But while it provides for our judgements of value an
object independent of our judgements, it fails to do justice to
what is also implied in our judgements of value, that when one
person says an object is good and another says it is not, they
are contradicting one another. For if M only meant 'I take an
interest in a' and N only meant 'I do not', they would not be
contradicting each other.
Professor Perry sometimes, for brevity, uses as equivalent
to 'good* "enjoyed by a subject*, and sometimes 'desired by a
subject'. Neither of these phrases does full justice to his theory.
His theory is that to be good is to be an object of interestj and
interest is thought of as covering both desire and enjoyment;
' Ib. 37. * Ib. 38.
i. e. the goodness of some things consists in their being enjoyed,
that of others in their being desired, and that of others, perhaps,
in their being both enjoyed and desired (though this, as we
shall see, is impossible). Now so long as we say (as he is apt to
say) that 'the goodness of the primrose consists in its being
desired',1 the theory seems at first sight attractive enough. But
obviously it is only a rough and ready description of my desire
to say I desire a primrose, What I desire is to be seeing it or
smelling it or possessing it As soon as we describe definitely
what it is that we desire, we see that it is something which does
not yet exist. There are no doubt cases in which we desire to go
on doing what we aredoing, or being in the same sort of state that
we are in. But even if I desire, for instance, to go on looking at a
primrose, what I desire is not the looking which is taking place
at present, but the looking which I wish to take place in the
immediate future. The object of desire is always something
non-existent. If it be said that it exists as a possibility, we must
reply that that is an inexact way of saying that the possibility of
it exists, which means that though it does not exist, the nature
of some or all of the things that do exist is not incompatible
with its coming into existence.
It is plain that in so far as goodness were either identical
with or dependent upon being desired, nothing could both
exist and be good. Now I suppose that we are all convinced
both that some things that exist now are good, and thatchings
of certain kinds, which may come into existence in the future,
will be good if and when they exist; and I suppose that apart
from these convictions we should have little or no interest to
the topic of 'good*, and ethics in particular would go by the
board. Yet in so far as the theory identifies the good with the
desired, it denies both these convictions. But it might be re-
plied that the goodness of existent things consists in their being
enjoyed, and the goodness of non-existent things in their being
desired. I must take leave, however, to doubt whether we can
say of a non-existent thing that it is good. However much one
were convinced that conscientiousness, for example, is good,
and that A might become conscientious, no one would say *^*s
Cf, A General Theory of f^alue, 133,
conscientiousness Is good* if he were convinced that A is not in
fact conscientious. But, our opponent might reply, we can say
of kinds of thing that they are good even if we are not convinced
that any instances of these kinds exist. We might say 'perfectly
conscientious action is good*, even if (as Kant suggests) we are
not convinced that there has ever been such an action. But
that is only a short-hand way of saying that without being sure
that such an action, ever has existed, we can be sure that if my
existed it would be good. Hypothetical goodness presupposes
hypothetical existence just as actual goodness presupposes actual
existence. And if so, being good can never be identical with
being desired, or even compatible with it,
The relation in which the primrose stands to desire is not
that of being desired but that of exciting desire. This is a rela-
tion in which existing things can stand to desire, and the theory
might be transmuted into the form, 'the good is that which
excites desire'. But the excitants of desire fall into two classes.
There are things our experience of which is such as to make us
desire to remain in our existing relation to them, or to get into
some closer relation to them, and to others like them; and there
are things our experience of which is such as to make us desire
to get away from them. Things of the second class are just as
decidedly excitants of some desires as things of the first class
are of others. And obviously one main sub-class (if not the
whole) of the second class consists of things that cause pain,
Thus if 'good' meant 'excitant of desire* we should be led to
the conclusion that things that cause pain are, as such, an im-
portant class of goods. This conclusion would evidently not
be accepted, and therefore the theory would have to be modified
into the form "the good is that which excites the desire to main-
tain our relations with it, or to get into closer relations with it,
and with others of its kind*—what we may, for short, call
'positive desire".
Now, on the face of it, some of the things that excite positive
desire1 do so because they are judged to be good. Prima facie
one would say that if the consciousness of a good disposition
Desire, of course, not for them to exist but for us to be in some new relation to
them or to continue to be in the same relation to them.
in oneself or the contemplation of it in another leads me to wish
to maintain and develop that sort of disposition, it is not because
I feel it to be pleasant but because I judge it to be good. But
this alternative is not open to Professor Perry, for, in basing
our taking an interest in the thing on our thinking it good, it
would involve the giving up of his main thesis, that a thing's
being good either is or is based upon our taking an interest in
it. All that is left for him therefore is to identify what is good
with that which by virtue of the pleasure it causes excites desire
for a closer relation with it and with other things like it. What
is good, then, for him is that which excites pleasure and thereby
excites such a desire. And though he includes both these ele-
ments in his formula, the fact of exciting pleasure is evidently
the root fact of which the other is a mere consequence.
Not only, however, is pleasantness the fundamental and
tendency to excite desire only a consequential element in good-
ness, according to the theory in the form in which it seems
necessary to restate it, but it is far more plausible to put forward
pleasantness, than to put forward this tendency, as the essence
of goodness. If we say 'that which produces so-and-so is, as
doing so, good", we are evidently implying that what is pro-
duced is intrinsically good, and what produces it instrumentally
good. And it is plausible enough to say 'pleasure is intrinsically
good, and what produces it instrumentally good'; there is a
pretty general agreement that pleasure, whether it is the good
or not, is at least good. But there is no general agreement that
desire, or even positive desire, is good. If we take the moral
standpoint we must say that some desires are good and others
bad, and that when desires are good they are good not because
they are desires but because they are the sort of desires they
are. And if we take the hedonistic standpoint, we must say
that desires are good or bad (which will mean "pleasant or
unpleasant') not in virtue of being desires but mainly (I suppose)
in virtue of their being supposed to be likely or unlikely to be
fulfilled. Desire (even positive desire) thus not being a thing
necessarily good in itself, there is no reason why, in general,
things that excite desire (or positive desire) should be good.
So long as we thought of things as objects of desire, it was
perhaps not implausible to say that objects of desire are good
even when the desire is not; but if we recast the theory in the
form in which we have found it necessary to recast it, and say
the good is that which excites positive desire, i, e. which is to it
as cause to effect, there is no reason (obvious or alleged) why,
positive desires not being always good, their excitants should
nevertheless always be so.
The most favourable way, then, of presenting the theory we
are examining is to exclude from it the reference to desire and
to reduce it to the form 'what is good is that which produces
pleasure'.1 But no one would in fact say that everything which
produces pleasure is good unless he thought pleasure itself
good;z and the theory emerges in the final form 'pleasure, and
pleasure alone, is good by its own nature j and what produces
pleasure, and only what produces pleasure, is good because it
produces something good.' The heart of the theory, then, in
spite of all it has said by way of attack on ordinary notions of
intrinsic good, is that there is one thing, and one thing only,
that is intrinsically good, viz. pleasure. The theory when re-
duced to its simple terms seems to be our old friend, hedonism.
After all the able refutations of hedonism that have been pub-
lished in recent years, it seems to me unnecessary to tread once
more on this rather hackneyed ground, and I suppose that
Professor Perry would agree that hedonism is untenable, and
claim that his own theory is tenable only in virtue of elements
that distinguish it from hedonism. But these elements are, if
I am not mistaken, among the least tenable elements in his
There is, however, one more point of view from which the
theory may be examined. Professor Perry describes 'the most
popular' objection to it as being that 'the fact of desire is not
accepted as final in most, judgments of value. Objects of desire
This is ambiguous, since it may mean * "good" means "productive of pleasure" ',
or 'what is good is good because it produces pleasure"; i.e. the ambiguity involved in
the theory from the start (cf, pp. So-i) still remains.
» And inferred from this that what produces it is good. But it is surely plain that
it does not follow from a thing's being good that what produces it is good, in the same
sense of 'good*. It roust be admitted that we often call "good* things that are merely
useful, but then "good* is being used improperly. Where I use the phrase 'instru-
raentally good", I use it to indicate this common but loose sense of 'good*.
are held to be bad in spite of their being desired, and desires
themselves are held to be bad whether or no they are satisfied'.1
I need not consider (a) one form of this objection with which
I have no sympathy—the view of Schopenhauer and others
that all desire is bad; that is an extravagance of quietism for
which there is little to be said. (3) The first real difficulty to
which the theory is exposed is that named next by Professor
Perry, viz. the fact that 'the same object may be liked or desired
by one man, and disliked or avoided by another',2 This fact,
taken with the identification of 'good* with 'object of interest*,
leads to the conclusion that the same thing may be both good
and bad. On the face of it, this result is paradoxical, and all
but self-contradictory; but he claims that 'a relational definition,
such as that here proposed, is the only means of avoiding con-
tradiction*.3 The claim is an odd one: by identifying good with
object of interest we get into the paradox of calling the same
thing good and bad (a paradox which an absolute theory at
least escapes, whatever be its other merits or demerits); and
then we triumphantly get out of the difficulty by saying, *Oh,
but good only means good for one person, and bad only means
bad for another person, so that there is no paradox.'
Is it not clear that when we assert the goodness of anything
we do assert something which we believe to be incompatible
with the same thing's being bad ? We may describe a thing as
"both good and bad1, but such language is not strict, (i) We
may mean that the thing contains some elements that are good
and some that are bad, but then that is the right way of putting
the matter, and 'the thing is both good and bad* is only a loose
way of putting it. It is implied in our thought on the subject
both that if we push our analysis far enough we shall find some
elements that are simply good and others that are simply bad,
and that the whole is not both good and bad but is either on
the whole good or on the whole bad. (ii) It may be suggested
that, without thinking of a thing as consisting of good and bad
elements, we may judge it to be good from one point of view
and bad from another—that a state of mind, say, may be morally
good and intellectually bad. But this turns out to be reducible
a 3
« A General Theory of Value, 154. Ib. 135. Jb. 136.
to the former case, in which analysis reveals a good and a bad
element. If we take a temporal section of the history of a mind,
however short be the section there will be elements in it of
knowledge and opinion which have a certain value, and actions
or dispositions to act which have a certain value (positive or
negative). The whole state of mind, then, cannot be judged
from the moral point of view, nor from the intellectual, but
some elements in it from the one and some from the other.
And each such element will have a goodness that is incompatible
with its being bad, or a badness that is incompatible with its
being good; and the whole state of mind will have a degree of
goodness or else a degree of badness, which can be assessed
only from a point of view in which we transcend both the moral
and the intellectual point of view.
(c) 'The case which has most deeply affected popular habits
of thought, and which is mainly responsible for the prejudice
against the present theory of value*, says Professor Perry, 'is
the case in which an interest or its object is morally condemned.*1
It is certainly an obvious objection to the theory that all objects
of interest are good, that in point of fact we do judge to be bad
many things in which nevertheless some one or other takes or
has taken an interest. Professor Perry's answer to this objec-
tion is to urge that in such a case we are performing a moral
judgement, and that 'moral judgments are not concerned with
value in the generic sense, but with a specific and complex
aspect of it.... They do not deal with interests per se> but with
the relation of interests to the complex purposes in which they
are incorporated.'2
In answer to this it is important to point out that the term
'moral judgement' contains a serious ambiguity. There are
three types of judgement which have by various writers been
termed moral judgements. These are (i) the judgements in
which an act is pronounced to be right or wrong; (ii) the judge-
ments in which an action or disposition is judged to be morally
good, or bad, or indifferent, L e. to have (or fail to have) the
kind of goodness or badness that only dispositions and actions
can have; (iii) the judgements in which something is said to be
1 2
Ik 136. Ib. 136-7.
good or bad ot indifferent sans phrase. The first two may
be -said to be departmental judgements, in the sense that each of
them is applicable only to one class of objects, the first to acts
considered apart from their motives, the second to dispositions
and actions considered in respect of their motives. Judgements
of the third class are not in any way departmental; they may be
made about anything in the whole world. It can be said of
some things-—I suggest, as at any rate an adumbration of the
things of which it can be said, virtue, knowledge and well-
grounded opinion, and pleasure1—that they are good; of
others—vice, badly grounded opinion, and pain—that they
are bad; and of other things that they are indifferent, I e» con-
sidered in themselves, though many ,of them may be instru-
mental to good or to evil. In making such judgements we are
not adopting a narrowly ethical standpoint; we are saying for
instance that wisdom and pleasure are good, though they are
not morally good. We are taking the most commanding
point of view that can be taken with regard to the value of the
things in the universe. Yet this is a point of view which a moral
philosopher should, in part of his inquiry, adopt, since ethics
is the study of that which we ought to do, and of what is in-
volved in its being what we ought to do, and since what we
ought to do depends to a large extent (though, as I have urged,
not entirely) on the goodness or the badness of the things we
can in our acts bring into being.
It would, however, be a mistake to spend time in arguing
the question whether the theory of good in general belongs to
ediics or to metaphysics. The other two types of judgement
belong exclusively to ethics. Goodness in general runs out
beyond the strict scope of ethics, if ethics be the philosophical
study of good conduct; for some of the things that are good
are neither conduct nor dispositions to conduct. But the study
of the meaning of good in general, and of the types of thing
that are good, is either a part of ethics, or a part of metaphysics
to which the study of purely ethical problems inevitably leads
us: which it is, depends on how we define ethics and meta-
To avoid making my statement too complicated, I omit a further kind of good
which will be mentioned later, cf, p. 138.
physics. Neither ethics nor metaphysics is a study to which
definite limits have hitherto been set, or one, probably, to
which they can profitably be set. The only way, perhaps, in
which we could prescribe a quite rigid programme for meta-
physics would be by saying that it is the study of the charac-
teristics possessed in common by everything that is; and from
this point of view the theory of goodness would have to be pro-
nounced not to be part of metaphysics. But whether we widen
our notion of metaphysics to make it include the theory of all
very widely distributed characteristics (among which goodness
and badness are included), or treat the study of value as a part
of ethics, or recognize an intermediate science of axiology (or
theory of value) less wide than metaphysics and wider than
ethics, is a question the discussion of which does not lead us
any distance at all towards understanding the facts.
We must return, however, to the objection Professor Perry
is at the moment considering, and to his answer to it. The
objection is that many of the things in which people find
pleasure and which they desire are nevertheless bad. His
answer is that they are not bad in general but only bad from
the ethical standpoint. And our answer to that is that while
there is what may be called a narrowly ethical standpoint from
which we judge such and such an action to be vicious or morally
bad, there is also a more commanding standpoint from which
we view the agent's total state of mind at the time and judge
that in spite of any elements of pleasure-value it may contain
it is on the whole a bad thing, a thing for whose occurrence
the world is the worse. This is not the narrowly ethical stand-
point, for it is the same standpoint from which we judge that
the occurrence of a pain is, considered apart from its accompani-
ments, a bad thing, though a pain is not morally bad. Now if
from this, which is the most commanding standpoint, we say
that many states of inind in which their owners have taken
interest and found pleasure are nevertheless bad, 'good' cannot
be identical with 'object of interest*.
My general conclusion is that Professor Perry's arguments
have not succeeded either in refuting the view that goodness
is an intrinsic quality of certain things, or in defending from
attack the view that it is identical with being an object of interest
to some mind.1
But the view that goodness is an intrinsic quality is exposed
to attack from another quarter. Professor Urban, whose work
on the question of values entitles him to the most serious con-
sideration, attacks this view and puts forward another,2 He
considers that there is a certain ambiguity in speaking of value
as a quality of objects. 'On the one hand, by value is often
understood such qualities as the good and the bad, the beautiful
and the ugly, the pleasant and unpleasant, etc. They are quali-
ties like yellow, hard, loud, etc. On the other hand, the value
quale is characterized, often by the same authors (Russell,
Meinong), as that which "ought to be on its own account" or
"as worthiness to be, or to be of interest".*3 The expression
value, he says, is not necessarily used "merely as a general term
for these qualities; it may also refer to the proposition that
the object ought to be, to something judged because of these
Professor Urban admits then that goodness is a quality, and
a quality like yellowness. But he thinks that besides the value
judgement 'so-and-so is good*, there is another type of value
judgement, 'so-and-so ought to be because it is good'; and it
is the latter type of judgement that leads him to reject a quality
theory of value. Here he seems to me both to admit too much
and to make a claim that cannot be sustained, (i) On the one
hand he admits that 'good*, 'beautiful*, 'pleasant* are qualities
like 'yellow", 'hard*, 'loud'. This seems to me to overlook two
important distinctions, (a) Even if 'good* and 'beautiful* are
qualities, they are not qualities like 'yellow* and 'hard*. There
is the vital difference—to be dwelt on later 5—that while the
latter are constitutive qualities, the former are resultant quali-
I may refer here to the weighty final chapter of Mcinong's last treatment of the
problem of value, in his Zur Grundkguag Jer allgemeinen Iferte&eerie, It is remarkable
that though he approaches the problem from the side of the subjective act of valuation,
and of the analysis of this, he concludes that there are "impersonal goods', in the sense
that there are goods which are not essentially for a subject at all, though they arc in a
subject (ef. p, 147 of his work). This is exactly the position I wish to establish,
Journal of Philosophy, &c,, 1916, 45^-65.
Ib. 4j6. * Ib. * Cf, pp. UI-*.
ties, depending on constitutive qualities, And (b) there is
a vital difference between "good* on the one hand and
'pleasant* or 'beautiful* on the other; for while it can be intelli-
gently asked whether the pleasant or beautiful has value, it
cannot be intelligently asked whether the good has value, since
to be good is just to be valuable, 'Good* is a more general
term than 'beautiful' and 'pleasant', and if we sometimes use
'good* with a narrower reference to moral good, it is with the
knowledge that "good* has a wider scope than this,
(2) But while he is content to regard not only 'pleasant* and
'beautiful* but 'good' as mere qualities. Professor Urban thinks
there is something further that may be asserted of things that
have these qualities, because they have them, viz, that they
ought to be on their own account, or that they are worthy to
be or to be of interest; and it is this that he identifies with value
proper, and asserts not to be a quality. This seems to me, with
all respect, to be a mistake, 'Ought* properly asserts an obliga-
tion, and it would be absurd to assert of, say, a state of pleasure
or a beautiful object that it is under an obligation to be, or of a
state of pain or an ugly object that it is under an obligation not
to be. It is true that we sometimes say of such things that they
ought or ought not to be, but it is always, if we use language
with any approach to strictness, with the underlying thought
that it is or was or will be some one's duty to bring them into
being, or to prevent them from being. It may further be pointed
out that it does not even follow from a thing's being good that
it ought to be, even in this loose sense, for if a thing, though
good,wouldadd less to the sumof good in the universe thansome
other good thing that can be produced by an alternative act,
it is (when no special obligations exist) our duty not to bring
it into being; and again there are special obligations which
over-ride the obligation to bring a pardcular good into being.
'This is worthy to be* or 'this is worthy to be an object
of interest* seems either (a) to be a metaphorical ascription of
moral merit, i. e. of a deserving to be brought into being or to
be made an object of interest, to things many of which (e. g.
states of pleasure or beautiful objects) obviously have no moral
merit; or else (£) to be a mere synonym for the phrase we have
already criticized—'this ought to be' or 'this ought to be of
interest'—and open to the same objection as these.
These two mistakes, as I venture to think them ( (i) and (2) )
are logically connected. It is because Professor Urban de-
grades 'good*, the universal adjective of value, to the level of
'pleasant* and 'beautiful', which are departmental adjectives of
value, and degrades all of these alike to the level of constitutive
qualities like 'yellow*, 'hard*, 'loud', that he feels a necessity to
seek value in something beyond goodness, and to express it in
the loose and metaphorical phrases he uses for it.
Further, he does not seem to be clear about the relation
between goodness, pleasantness, or beauty, and 'ought-to-be-
ness'. For on one pagel he describes the proposition that an
object ought to be as 'something judged because of these
qualities', and on the next he says that from the presence of
these qualities 'the positive value of the object itself does not
follow*. Why then should one judge that it ought to be, "because
of these qualities* ?
Professor Urban is seeking to show that value, like existence,
is not a quality, and states this in the following way. 'As an
object may have its full quota of qualities and the question of
its existence or non-existence still be left open, so an object may
have its full quota of qualities, including its so-called value
qualities, and we may still have to ask whether it ought to be
or not.*2 Here he seems to be on the track of a true distinction,
which, however, I should prefer to state in another way. There
are a number of characteristics which evoke some sort of ap-
proval or preference and in virtue of which anything that has
them has some intrinsic goodness. Yet to say of a thing that it
has one of these is not to say that it is intrinsically good. Treat-
ing beauty as a form of instrumental goodness (i. e. not as a
form of goodness in the strict sense but as a way of being pro-
ductive of something good),3 I do not reckon it among these
characteristics. The three which seem to me to cover the
ground (though I may easily be mistaken) are the character-
istics of being virtuous, of being an instance of knowing, of
1 2
Journal of PUtesopfyf&x., 456. Ib. 457.
» Cf. pp. ii?-ji.
being an instance of being pleased.1 If therefore a state of mind
possessed one of these valuable characteristics and did not
possess any of the opposite characteristics, we should neces-
sarily judge it to be good. But concrete states of mind, perhaps
always and certainly sometimes, occupy positions on all three
scales—that of virtue-vice, that of inteUigence-unintelligence,
that of pleasure-pain; and it therefore does not follow, because
a state of mind has one, or two, of the characteristics which in
themselves are good, that it is good on the whole. Thus to
judge of a state of mind (or of a state of society) that it is on
the whole virtuous, or intelligent, or pleasant, is not necessarily
to judge that it is good on the whole. The judgement that a
state of mind is good on the whole is what I should put in place
of Professor Urban's judgement that a thing ought to be, as
distinguished from the judgement that it is virtuous or pleasant
or beautiful. And if I am right in holding that 'this is good* is
the true expression of this ultimate valuation, rather than 'this
ought to be*, the argument for the view that goodness is not a
quality, and for the assertion that its affinities are rather with
existence, falls to the ground. I would distinguish as Professor
Urban does between the possession of moral goodness or of
intellectual goodness or of feeling-goodness by a thing (i.e. by
a state of mind) and its possession of goodness on the whole.
But I see no ground for his view that while these departmental
values are qualities, value 'is not a "what" at all, either quality
or relation: it is a "that" *. Is it not clear that value must fall
within the same category as its own varieties ?
Professor Urban admits this as a possible view, i. e. that
ultimate value may be a quality, though different from the
'valuable qualities'.2 But he holds that this view has been
refuted by Croce. Croce's argument,3 as summarized by Pro-
fessor Urban, is this: 'Take . . . the value-judgement in its
usual form, A is as it should be, or negatively, A is as it should
not be. The first, he holds, is tautology, the second a logical
absurdity. If A exists, it is already as it should be, for it can
not be other than it is.*4 The conclusion Croce draws is that
This last requires some modification ;cf, pp. JJ 5—8. * Journal of Philosophy,
&c., 459. * Seggio suits Hegel, 409. * Journal of Philosophy, &c,, 459.
the so-called value judgement is not a judgement at all but a
mere expression of feeling. Professor Urban draws a different
conclusion—that value is not a quality. But let us see if either
of these conclusions really follows. It is necessary to dis-
tinguish between two types of subject of which it might con-
ceivably be asserted that it is (or is not) as it should be. (i) It
might conceivably be said 'pleasure is as it should be*, 'pain is
not as it should be*. And against any such judgement Grace's
criticism is justified. To say for instance that pleasure is as it
should be implies that pleasure, while remaining pleasure,
might have different natures, that it is better that it should have
one of them than any other, and that it has that which it is best
that it should have. And so with any similar judgement. Of all
such judgements we may say, not with Croce that some are
tautological and others are absurd, but that all are absurd. For
in naming the subjects of our judgements as we have named
them, we have ascribed to them a certain nature and excluded
the possibility of their having a different nature, and also (since
the value of things depends on their nature) the possibility of
their having a different value, from that which they have. This
is just what distinguishes intrinsic from instrumental value;
for if a thing is only instruraentally good or bad, then even
when its nature remains the same it might have a different
instrumental value if the causal laws of the universe, or the
other things in the universe, were different.
But it is to be observed that, so far, Croce is fighting a man
of straw. No one can judge that 'pleasure is as it should be*
unless he has adopted (as Croce does) the artificial view that
A is as it should be' is the correct form of the value judgement.
If we adopt instead the natural form, 'pleasure is good', 'pain
is bad', his argument leaves us quite untouched. For to point
out that pleasure or pain could not be other than it is, has not
the slightest tendency to show that the judgement 'pleasure is
good* is tautological or the judgement 'pain is bad* absurd.
(2) There is another and quite different type of judgement, in
which we can and do assert of something that it is or is not as
it should be. We may say "the state of so-and-so's mind is not
as it should be*, or 'the state of England is not as it should be'.
Here we do not in naming our subject already describe its
nature and exclude the possibility of its having a different
nature, and a different value, from that which it has. We mean
that the existing state of things is bad, with the further implica-
tion that it is the duty of some person or persons to substitute
for it a different state which would be good. And this again is
left entirely untouched by Croce's argument.
The supposed collusiveness, then, of Croce's argument
against value being a quality arises partly from the use of the
phrase 'A is as it should be' as the true expression of the value
judgement, and partly from a confusion between a kind of
judgement of this form which is absurd but which no one actu-
ally makes, and another type of judgement of this form which
we do make but which is not tautological nor absurd.
It is possible to exhibit in yet another way the falsity of
Professor Urban's view that the affinity of value is with exis-
tence and not with qualities, "Qualities inhere in objects,* he
says, 'since it is the quality that makes the object precisely what
it is; the judgement of quality presupposes that the thing is not
other than it is. But the object may have its full quota of quali-
ties without being judged valuable any more than existent,* *
He is thinking, no doubt, of Kant's famous criticism of the
ontologica] argument on the ground that it treats existence as
a quality among others. An imaginary thaler may have (or
rather be imagined to have) exactly the same qualities as an
actual thaler in my pocket; existence is not one of the qualities
constitutive of a thing's nature, nor something that follows
from them. But surely the analogy entirely breaks down. For
whereas from the fact that an imaginary thaler is imagined to
have exactly the same qualities as a real thaler has, it does not
follow that since the one exists the other exists, from two things'
having exactly the same qualities other than goodness, it most
decidedly does follow that if the one is good the other is good.
While existence is not a quality, goodness is a quality conse-
quent on the other qualities of that which has it.
Value is, Professor Urban maintains, "not an adjectival
predicate, but an attributive predicate,... It is a predicate only
Journal af PAikmpfty, &c., 459,
in the sense that existence and truth are predicates';1 i.e. he
considers that there is a class of judgements of which *A exists*,
'A is good', 'A is true* are instances, in which no quality or
relation is ascribed to a subject. And though, as I have tried to
show, there is a quite essential difference between our predica-
tions of existence and our predications of value, to which his
view does not do justice, it is perhaps worth while also to
attack the view that there is a class of non-qualitative, non-
relational predicates, of which 'existent* is merely one. The
judgement fA is true* is plainly a judgement in which another
judgement is asserted to possess a certain quality, which in turn
depends on a certain relation between the judgement that is
judged about and an existing fact. It is not of the same type as
the existential judgement, and there is no argument from analogy
for supposing that the judgment SA is good* is so,
In contrast with the views that value is a relation, or a quality,
Professor Urban inclines to the view that value is 'an objective,
or specific form of objectivity',2 and he selects three points as
sufficient to characterize this view, *(i) Value is ultimately
indefinable in the terms or categories of matter of fact—as
object, or quality, or relation; (2) the judgment of intrinsic
value, that an object ought to be, or to be so-and-so, on its own
account apprehends an ultimate and irreducible aspect of ob-
jects; (3) this value is itself, not a quale of some objects, but is
a form of objectivity, in contrast with being and existence,'3
I may be allowed to sum up my own attitude towards the
theory as follows, (r) If 'object' here means, as I suppose it
must, 'substance', value is certainly not that. And I have tried
to give reasons for holding that it is not a relation. But the
arguments against the prima fade view that it is a quality seem
to have completely broken down, (2) I agree that the judge-
ment of intrinsic value apprehends an ultimate and irreducible
aspect of objects, and think that this is the sole element of truth
—and a very important element of truth—in the view, (3) I
reject the view that the true form of the value judgement is not
'A is good' but 'A ought to be' or 'A is as it ought to be', which
if accepted might lend some colour to the theory that the value
Journal of Philosophy, &c«, 459-60. * It>. 461. > Ib,
judgement is akin to the existential judgement 'Ah' rather than
to the predicative judgement 'A is B*,
Under the third head, Professor Urban recognizes two types
of the theory, one which holds that value is what ought to be,
another which holds that it is what ought to be acknowledged,1
The second form, is no more acceptable than the first; for the
only conceivable ground for holding that a thing ought to be
acknowledged, i. e. acknowledged to exist, is that it exists, so
that the fact that it ought to be acknowledged is not the funda-
mental fact about it but only a consequential one.
Professor Urban adds two points which he considers to tell
in favour of his view, (i) The first consists in emphasizing the
verbal form of the value judgement. 'In the value judgment',
he insists passim,2 'we apprehend a "that", not a "what".' And
the ground here given is that what we judge is not an object
but an objective, i. e. something which can only be expressed by
a thai-construction. Now we may if we please call anything of
which the proper expression is a that-clause an objective; I
will accept this language for the moment. And if so we must
agree that what is judged in a value judgement is an objective.
But how does this prove that value is an objective ? For it is
equally true of any judgement that what is judged in it is an
objective. But the fact that 'that A is red* is an objective does
not show red to be an objective and not a quality; how then
does the fact that 'that A is good* is an objective show good to
be an objective and not a quality ? In this respect the two judge-
ments are precisely parallel.
But while value appears not to be an objective, I am inclined
to think that it is only objectives that in the long run have value,
or that have ultimate value. This fact, if it be a fact (the ques-
tion seems to me a very difficult one), is obscured by certain
perfectly natural and not really misleading but nevertheless not
strictly accurate ways of speaking. We say for instance of a
man that he is good. But if we ask ourselves what is really good
in or about him, we find that it is not all that he is or all that he
does, but those actions, dispositions, or states of him which
have a certain character; his acting, say, on certain occasions
Journal of Philotoplgr, &c,, 461. * e. g, ib, 461.
from love of duty. Now that which is valued, like that which
is desired, is not so obviously an objective as that which is judged,
For the latter we have only one proper form of expression.
No one ever says *I judge his goodness*; we have to say *I judge
that he is good*. For the object of valuation, and for the object
of desire, we have other forms of expression. We can say 'his
acting from such and such a motive is good', or 'his insight into
a comprehensive law of the universe is good", or 'his being
pleased is good*; and we can say 'he desires the happiness of
his children*. But in both cases we can quite naturally substitute
the thai-form. We can say 'that he is acting from a sense of
duty is good', or 'that he has insight, &c, is good', or 'that he
is experiencing pleasure is good5; and in the case of desire we
can say 'he desires that his children shall be happy*. Thus the
that-construction seems to be a comprehensive form applicable
to what we value and to what we desire, as much as to what we
judge, though in the former two cases language has provided
us with alternative expressions which mean the same as the
'that* clause.
This, if true, is important; for it enables us, without denying
goodness to be a quality, to recognize a great difference between
it and most qualities. Most of our adjectives, 1 suppose, refer
to qualities that belong to substances; 'good* is the name of a
quality which attaches, quite directly, only to 'objectives', and
since an objective is an entity more complex than a substance,
standing as it does for a substance's having a certain quality or
being in a certain relation, 'good* may be called a quality of a
different type from those that attach to substances. To this it
might be objected that I judge that so-and-so's- acting conscien-
tiously, or knowing an important law of the universe, or being
pleased, is good, simply because I judge that virtue, knowledge,
or pleasure is good. But it is surely clear that 'virtue', 'know-
ledge*, and 'pleasure* are just short-hand ways of referring to
the fact that some mind or other is in a certain condition, i. e. to
entities that are complex as contrasted with mere substances,
or mere qualities, or mere relations; in other words, to ob-
But if I use the word 'objective* as a convenient way of
referring to the things that can be expressed by the *dass-
construction', by 'that'-clauses, I am far from thinking that
these taken together form a class of real entities. If I think 'that
A is B\ this does not imply that there is a real entity 'that A is
B'—an independently real proposition with which my mind
is in a certain relation. If I say 1 think that A is B\ It is my
state of mind and not any independently existing entity that I
am describing. The reasons which have led some logicians to
recognize propositions as entities distinct both from objective
facts and from the activity of thinking appear to be false; as the
subject is foreign to our inquiry, I am content to refer to Mr.
McTaggart's refutation of them.1 And similarly, if I desire
'that A shall be ff, this does not imply that 'that A shall be B'
is a real entity. In saying 'I desire that A shall be B' I am de-
scribing the state of niy mind and nothing else; there is no
entity 'that A shall be B' with which my mind enters into rela-
tion. Thus the whole variety of things that are expressed by
'that'-clauses do not form a class of entities. But some of the
things so expressed are real entities. If I know 'that A is B\ it
is implied that 'that A is B' is a real element in the nature of the
universe. And if'that mind A is in state B' is good, then again
it is implied that 'that mind A is in state B' is a real element in
the nature of the universe. But the proper name for what is
expressed in such 'that'-clauses, which are real elements in the
nature of the universe, is not 'objectives' (which suggests an
affinity that does not really exist between these and what is
expressed in the other 'that'-clauses mentioned above), but
"facts'. It is better therefore to say that the things that have
ultimate value are facts. And since these are entities of a higher
order of complexity than substances, we get an important
distinction between value judgements and the judgements in
which we judge about substances,
(2) Professor Urban argues that 'in order to know that an
object ought to be or is as it ought to be, it is not necessary to
know whether the object is or is not'.2 This statement, in the
second of its alternative forms, is plainly not acceptable. It is
clearly impossible to know that anything *is as it ought to be*
' The Nature t»f Exutenee, i. $s~3*. * Journal of Philosophy, &c., 463.
without knowing whether it is at all. The other form, 'such
and such an object ought to be", contains no such obvious pre-
supposition of the object's existence. But we have already
given reasons for rejecting this as a correct expression of the
value judgement. The true expression of that judgement is 'A
is good*. Does this presuppose A*?, existence? Professor Urban
maintains that it does not, and this is, I think, involved in his
view that value is 'a form of objectivity, in contrast with being
and existence*,1 'lying between being and non-being, but itself
not a form of being*.2 He maintains that such a thing as perfect
happiness can be judged to be good, without being known to
be possible, let alone existent. But it seems clear that if we do
not judge perfect happiness to exist, the proper form of ex-
pression ts not 'perfect happiness is good' but 'perfect happiness
would be good*, i. e. if it existed—where its being good pre-
supposes its existence; just as it is the case that, if we do not
judge that A is happy, we cannot say 'A'?, happiness is good*
but only *A\ happiness would be good*. The judgement that
a thing is good presupposes the judgement that it exists; and
the judgement that it would be good presupposes the sup-
position of its existence. Wecan, of course, make the judgement
that it would be good if it existed, without knowing or even
judging that it exists, or even that it is possible. But that in no
way tends to show that value is independent of existence.
Actual value presupposes actual existence, and conditional
value supposed existence.

I conclude, then, that the arguments in favour of thinking

of value as an objective are no more successful than those in
favour of treating k as a relation, and that the arguments against
its being a quality put forward by those who hold the *objec-
tive'-theory are no more valid than those put forward by the
advocates of the 'relation'-theory. The natural view that value
is a quality therefore holds its ground, and we may proceed to
consider whether we can say anything about the sort of quality
it is, and its relation to other qualities. In considering this
I have been much helped by Professor Moore's study of the
' Journal ef Philosophy, &c., 461. Ib, 464.
conception of intrinsic value,1 the main points of which I will
briefly summarize. 'To say that a kind of value is intrinsic is to
say that the question how far a thing possesses it depends
entirely on the intrinsic nature of the thing, i. e, (i) that it is
impossible for a self-same thing to possess the value at one
time and not at another, or to possess it more at one time and
less at another; (2) that if a thing possesses any kind of intrinsic
value in a certain degree, anything exactly like it must in all
circumstances possess it in the same degree. Or to put it other-
wise, if of two things not having a different intrinsic nature
one has a certain value, the other must also have it.
'Intrinsic difference is not qualitative difference (though
most intrinsic difference is qualitative difference). For things
that possess the same quality in different degrees are intrinsi-
cally different, e. g. a loud sound and a soft one, or two things
of different sizes; and so are a yellow circle with a red centre and
a yellow circle with a blue centre, although not these wholes
but only single elements of each are qualitatively different.
'What is meant by "impossible" when we say "a kind of value
is intrinsic if and only if it is impossible that X and Y should
have different values of this kind unless they differ in intrinsic
nature"? (i) It is sometimes said that "it is impossible for that
which has attribute F to have attribute G" means simply that
things that possess F never in fact possess G. But more is
meant than this.2
*(ii) There are causal laws in virtue of knowing which we
seem able to say that if a thing had a property F it would have
a property G. But when we say that if a thing has a value in
a certain degree, anything exactly like it would have the value
in the same degree, we mean that this is so independently of
any causal laws.
'It is clear that there is such a sense of necessity. If you take
a particular patch of colour, it is certain that any patch exactly
like it must in any universe be yellow. It is in a similar sense
that we assert that two things exactly alike must have the same
Philosophical Stu&s, 255-75.
I omit ProfessorMoore'sreasont for this assertion, which i*(lthink)ol»vkjusly true.
*(tii) The necessity we are examining is, finally, not identical
with logical necessity, such as that which we assert when we
say that whatever is a right-angled triangle must be a triangle,
"I do not see how it can be deduced from any logical law that,
if a given patch of colour be yellow, then any patch which were
exactly like the first were yellow too." So too it cannot be
deduced from any logical law that if A is beautiful anything
exactly like A would have the same degree of beauty.
'But while both yellowness and beauty depend only on the
intrinsic nature of what possesses them, yellowness is and
beauty is not an intrinsic predicate. No predicate of value is an
intrinsic predicate in the sense in. which yellowness, or being
a state of pleasure, is intrinsic. Apparently certain predicates
of value are the only attributes that share with intrinsic proper-
ties the characteristic of depending solely on the intrinsic nature
of what possesses them. The fact that many people have
thought that values were subjective, points to some great
difference between them and such attributes as yellow. What
is the difference, if values are not subjective ? It may be vaguely
expressed by saying that intrinsic properties describe the nature
of what possesses them, in a way in which values do not. If you
could enumerate all its intrinsic properties you would have
given a complete description of the object. It is true both of
intrinsic properties and of values that if one thing possess them
and another do not, the intrinsic nature of the two things must
be different. Is the necessity of a different kind in the two
cases? If it is, we have two senses of "must", both uncon-
ditional, and both different from the logical "must".'
This analysis seems to me a valuable one. I agree both with
Professor Moore's account of what is implied in a value's being
intrinsic, and with what I take to be his view, that goodness,
in its most fundamental meaning, the meaning to which I have
tried to call attention by eliminating the 'good of its kind* and
the 'instrumentally good', is intrinsic. But I wish to express a
difference of opinion with regard to Tightness and beauty, the
two attributes which he describes as the main value-attributes
other than goodness; to examine further the way in which the
goodness of a thing depends on its intrinsic nature; and to
comment on some points of detail. I begin with the latter,
(i) The first concerns the distinction which Professor Moore
draws between difference in intrinsic nature and difference in
quality. He holds that there are things that differ in intrinsic
nature without differing in quality, and he gives two types of
instance of this, (a) 'Nobody would say that a very loud sound
was exactly like a very soft one, even if they were exactly like
in quality; and yet it is plain there is a sense in which their
intrinsic nature is different*1 Similarly no one would hold
that a particular patch of colour was "exactly like' another patch
of identical quality but a million times larger or smaller, or that
if the one were beautiful the other must have exactly the same
degree of beauty. On this I would observe that if loudness and
size are not qualities (a question on which I offer no opinion),
the point will be met by saying that the intrinsic nature of a
thing includes besides its quality whatever quantity (intensive
or extensive) it may have.
(f) The second illustration of difference in intrinsic nature
as distinct from difference in quality is "the difference between
two patches consisting in the fact that the one is a yellow circle
with a red spot in the middle, and the other a yellow circle with
a blue spot in the middle*.2 This, which is obviously a differ-
ence in the intrinsic nature of the two patches, is described
by Professor Moore as only loosely to be called a difference
of quality. But surely he holds that two things not different in
quality are identical in quality, and surely no one would ascribe
identity in quality to two such patches. The expression 'differ-
ence in quality* seems quite sufficient to cover such a case as
this, if only we recognize that 'different in quality' means
'having any difference in quality', not necessarily 'different
throughout in quality*.
It would seem then as if we might say that the intrinsic nature
of a thing includes its quality and its quantity. But the question
may naturally be asked, should not relations also be brought
into our account of the intrinsic nature of a thing? Obviously
the relations of a thing to things other than it and other than
its elements should not. For the very view that a thing has an
* PAiltHopliieal StuJctt, 164. * Ib.
intrinsic nature is the view that it has a nature which does not
include any relations of it to other things.1 There remain,
however, (a) the relations of the thing to its own parts or ele-
ments, and (&) the relations between its elements. Must these
be included in its intrinsic nature as separate factors, over and
above its quality and its quantity? It will be agreed that two
patches of colour have not the same nature if one consists of
three bands of red, white, and blue, in that order, and the other
of red, blue, and white, in that order j and that if one was beauti-
ful it would not follow that the other was. And this difference
of intrinsic nature rests on a difference not between the quality
or quantity, but between the relations, of the elements of the
two wholes. Nevertheless it seems clear that it is part of the
quality of the one wkok to consist of the three bands in one
order, and part of the quality of the other whole to consist of
the three bands in a different order. And if this be so, the rela-
tions of the parts need not be included in the intrinsic nature of
the whole, over and above the quality of the whole. And for a
similar reason the relations of the whole to the parts need not
be included as a separate factor. For if one patch contained a
yellow part which was half the size of the whole, and another
a yellow patch which was a third of the size of the whole, these
facts would be part of the quality of the wholes. It would
appear, then, as if intrinsic nature included simply quality and
(2) Another point which invites comment is the nature of
the necessity or necessities by which (a) if one patch of colour
is yellow, another having the same intrinsic nature must be
yellow, and {b) if one patch is beautiful, another having the
same intrinsic nature must be beautiful. I agree with Professor
Moore in thinking that in neither case does the necessity mean
(i) that in fact no two patches of colour of identical nature are
respectively yellow and not yellow, or respectively beautiful
and not beautiful. This will be true, of course, but we mean
more than this. Nor does it mean (ii) that in this universe there
is a causal law which ensures that no two patches of colour
' Cf. Cook Wilson's analysis of the meaning of "what a thing is in itself', Statement
and Inference, i. 15*-}.
intrinsically alike are respectively yellow and not yellow, or
respectively beautiful and not beautiful. I think that we can
see that this must be so in any universe, and cannot depend on
the causal laws of this universe. When he asserts (iii) that the
necessity is not logical necessity, the question becomes more
difficult. He describes logical necessity as 'the kind of necessity,
which we assert to hold, for instance, when we say that what-
ever is a right-angled triangle must be a triangle, or that what-
ever is yellow must be either yellow or blue*, * I agree that from
the single proposition "these two patches of colour are intrinsi-
cally alike* it does not follow that if one is yellow the other
must be yellow, or that if one is beautiful the other must be
beautiful But elsewhere he gives, and is, I think, evidently
right in giving, as a further instance of logical necessity, the
fact that 'from the proposition with regard to any term that it
is red it follows that it is coloured*.2 Here I suppose the con-
clusion follows strictly not from the one proposition 'this is
red* but from this plus the self-evident proposition that red
is a colour. I suggest that the necessity by which if two
patches of colour are intrinsically alike it follows that if one
is yellow the other is yellow, is of the same nature; L e. this
follows not from the bare proposition 'these two patches
of colour are intrinsically alike*, but from this plus the self-
evident propositions that yellow is a quality (which is of the
same type as 'that red is a colour' above), and that intrinsic
likeness involves identity in quality. The necessity is what
Professor Moore calls a logical necessity; I should prefer to
say that it is an instance of implication and not of causality.
What are we to say of the necessity by which if two things
are exactly alike it follows that if one is beautiful the other is
beautiful ? This question is perhaps better left till we examine
the difference which Professor Moore now points out to exist
between beauty and such attributes as yellowness. 'The differ-
ence I mean is one which I am inclined to express by saying
that though both yellowness and beauty are predicates which
depend only on the intrinsic nature of what possesses them, yet
while yellowness is itself an intrinsic predicate, beauty is not.'
' PlultuofiucaiStudies, tji. Ib, 384-^.
He suggests that intrinsic kinds of value, if any exist, are unique
in respect of the fact that, though not intrinsic properties, they
'share with intrinsic properties this characteristic of depending
only on the intrinsic nature of what possesses them'.1 He con-
fesses himself unable to define further the difference between
value and intrinsic properties except 'by saying that intrinsic
properties seem to describe the intrinsic character of what
possesses them in a sense in which predicates of value never do.
If you could enumerate all the intrinsic properties a given
thing possessed, you would have given a complete description
of it, and would not need to mention any predicate of value it
possessed; whereas no description of a given thing would be
complete which omitted any intrinsic property'.2
Here the use of the terms 'depend* and 'complete descrip-
tion' seems to me open to criticism. To say that both intrinsic
properties and value depend on the nature of their possessors
is to suggest that it has an intrinsic nature apart from both and
on which both are based. But Professor Moore's conception,
and the true conception, of intrinsic properties is that they are
just the things that make up the intrinsic nature of their pos-
sessors. What are the minimum conditions that must be ful-
filled if two patches of colour are to have the same intrinsic
nature ? Let us say that they are that the two patches must be
(i) of the same colour-tone, (2) of the same brightness, (3) of
the same size, (4) of the same shape. Any of these may vary
independently of any and all of the others. The essential differ-
ence between them and beauty is that beauty, and the degree
of beauty, cannot vary independently of them taken together.
Two patches of colour exactly alike in these other respects
must be alike in beauty. In other words, while these others are
constitutive characteristics, beauty is, in some way (which we
must later 3 try to describe more precisely), a dependent or
consequential characteristic. The necessity that if the intrinsic
nature of the two patches is the same their colour (or brightness,
or size, or shape) must be the same is analytic; the necessity
that their degree of beauty must be the same is synthetic.
Again, I cannot agree that a description of a patch of colour
' Philosophical Studies, 272. * It>, 174, > Cf. pp. 1*1—3.
would be complete without the statement that it is beautiful (If
it is so); for its beauty might be for some purposes the most
important fact about it. But it would be, in a certain respect
(i. e, for a certain purpose), sufficient', in that respect in which a
description of a triangle as being an equilateral triangle with
sides a foot long would be sufficient. Any such triangle would
have many more attributes; it would be equiangular, its angles
would be angles of sixty degrees, it would have a certain area,
&c.j but a description that omitted such attributes but men-
tioned its equilaterality and the length of its sides (or of one
side) would be sufficient in the sense that two triangles alike in
these respects must be alike in all respects, and that it would
state the attributes from which all the other geometrical attri-
butes of such a triangle necessarily follow. But it would not be
sufficient in respect of telling us all that the triangle is; and no
more would be the description of a patch of colour which
omitted to say whether it is beautiful or not.
In fact, as a first approximation at any rate, we may say that
the difference between goodness or value and such attributes as
yellowness is that whereas the latter are differentiae (i. e. funda-
mental or constitutive attributes) of their possessors, the former
is a property (i, e, a consequential attribute) of them. And in
this respect goodness may be compared to the properties that
geometry proves to hold good of various types of figure. But
if we recognize the affinity, we must also recognize the marked
difference between the two cases, (i) In the first place, it is
quite arbitrary which of the attributes equilaterality and
equiangularity is selected as the differentia of the kind of
triangle which in fact possesses both; it is just as proper to
say that the relative size of the angles determines the relative
length of the sides, as vice versa. Value, on the other hand,
seems quite definitely to be based on certain other qualities of
its possessors, and not the other qualities on the value. In fact
the distinction between differentia and property as fundamental
and consequential attributes respectively is far more appro-
priate in this case than in most of those to which it is applied,
since this is one of the comparatively few cases in which one of
the attributes is objectively fundamental and the other objec-
lively consequential. We might call value a genuine as opposed
to an arbitrarily chosen resultant (or property).
(2) Another respect in which value differs from mathematical
properties, as well as from certain other resultant properties,
is that while mathematical (i. e. spatial, temporal, and numerical)
properties follow from part of the intrinsic nature of their pos-
sessors, value follows from the whole intrinsic nature of its
possessors. If a patch of colour is in shape an equilateral tri-
angle, it will be an equiangular triangle, whatever be its size or
colour^ if it is of a certain colour, it will match certain other
patches,, whatever be its shape or size; if it is of a certain size, it
will be equal in size to certain other patches, whatever be its colour
or shape. These attributes which are based on some single
element in the nature of their possessors may be called parti-
resultant properties. In contrast with these, value is a toti-
resultant property, based on the whole nature of its possessors.
And this is true not only of "good*, the adjective which ex-
presses intrinsic value, but also of'right' and 'beautiful', which
are often classed with it; though "right' does not stand for a
form of value at all,1 and 'beautiful* does not stand for an
intrinsic form of value.
(a) Take first the attribute 'good* in the sense we are examin-
ing, i e. that in which it means 'intrinsically good*. If I may be
allowed to anticipate positions which I shall try to commend
independently of the present argument, good is a characteristic
belonging primarily only to states of mind, and belonging to
them in virtue of three characteristics—the moral virtue
included in them, the intelligence included in them, and
the pleasure included in them.2 The precise degree of good-
ness belonging to any state of mind will belong to it not in
virtue of any one or two of these elements but in respect of all
of them, or of all which it possesses. Two states of mind alike
in respect of virtue and intelligence are not equally good if
they differ in their degree of pleasantness; nor two states
alike in virtue and pleasantness equally good if they differ in
their degree of intelligence; nor two states alike in intelligence
What has value is the doing of the right act because it is right,
* Of. a, i on p, 107.
and pleasantness equally good if they differ in their degree of
(£) An act that is right is right in virtue of its whole intrinsic
nature1 and not of any part of it In respect of certain elements
in its nature it may be prima facie right, and in respect of others
prima, facie wrong j whether it is actually right or wrong, and
if it is wrong the degree of its wrongness, are determined only
by its whole nature. This point has been, I hope, sufficiently
dealt with in our second chapter.
(c) It is, I suppose, still more clear that it is only in virtue of
its whole intrinsic nature that any beautiful thing has the precise
degree of beauty that it possesses. Take a simple case. A patch
of colour might have been more, or less, beautiful than it is if
either its shade of colour or its degree of brightness or its size
or its shape had been different. Or, to take a more complicated
case, the alteration of a poem in any syllable may, in principle,
make it more, or less, beautiful.
If we are agreed that goodness, rightness, and beauty depend
on nothing less than the whole intrinsic nature of their pos-
sessors, we must consider whether they depend on anything
more than this. Only if they do not, only if we can say 'if A is
good, right, or beautiful, anything exactly like A will be equally
good, right, or beautiful*, will goodness, rightness, and beauty
be intrinsic properties. An apparent objection naturally arises
in the case of beauty. There might be two patches of colour
A, A' exactly alike. Yet if one lay beside a patch of colour B
which harmonized with it, and the other beside a patch C which
clashed with it, it would seem natural to say that the one might
be beautiful and the other ugly. The objection is, however,
illusory. For strictly it is not the two like patches A, A' that
would be respectively beautiful and ugly, but the wholes A B,
A' C, On reflection we should admit that though A B is beauti-
ful and A' Cugly, still HA'is beautiful, A' if exactly like it must
be beautiful too; though to apprehend its beauty we should
have to do what it may be difficult to do, viz. contemplate it in
abstraction from C,
i, e. in virtue of being the origination of aE that it is the origination of. But ef. the
qualification in n. % p, jj.
It is, I think, clear that the only types of theory which deny
that beauty is intrinsic to beautiful objects are those which
make it consist in or depend on the fact that some mind or
minds do or undergo something in face of them. And such
theories, as we have seen,1 fall into two main kinds, according
as beauty is held to depend on the occurrence of (A) a feeling
or (S) a judgement. We have seen, too, that such theories may
connect beauty with the occurrence of some feeling or judgement
(i) in at least one person, no matter in whom, (2) in the person
who judges the object beautiful, (3) in a majority of persons
of some class or other—say aesthetically cultivated people, or
people belonging to a particular stage in the history of civiliza-
tion, (4) in a majority of mankind, or (5) in all mankind. The
most appropriate names for the feeling in question are, I
suppose, 'aesthetic enjoyment* or 'aesthetic thrill*, and for the
opposite feeling 'aesthetic repulsion'. Into the precise nature of
these feelings it is not necessary for our argument to inquire.
It is perhaps enough to say that we all know these feelings by
experience, and have a general idea of what they are like.
Corresponding theories, we have seen,2 may be held with
regard to 'right* and with regard to 'good*. Professor Moore
rejects theory (A i) with regard to 'right* on the ground (a)
that *a man may doubt whether an action is right, even when
he does not doubt that some man or other has the required feeling
towards it*.3 The argument seems to me conclusive in the case
of 'right"; but a corresponding argument is not equally con-
clusive in the case of'beautiful*. For it must surely be admitted
that the nature of aesthetic enjoyment and of aesthetic repulsion
is more obscure, and more liabie to be confused with something
different, than is that of moral approval or disapproval. A man
who thinks that he himself gets aesthetic enjoyment from a
certain object may well doubt whether any one has ever ex-
perienced genuine aesthetic repulsion to it. He may hold that
what the other feels is a non-aesthetic repulsion to it arising
from, e. g., moral or religious or patriotic motives, or from the
shock of novelty, though the other may wrongly think and
say that his repulsion is aesthetic. And similarly a man who
1 2 J
p. So. pp. n, 80. Ethks, in.
thinks that he himself experiences aesthetic repulsion to an
object may well think that no one has ever experienced genuine
aesthetic enjoyment from it, and that those who thought they
did were mistaking a pleasant feeling of some other kind (moral
approval, or religious enthusiasm, or the like) for the genuine
aesthetic thrill. Is not this, in fact, the account we are most
likely to give when we find ourselves differing from other
people about the beauty or ugliness of an object? It is not,
therefore, clear that *a man may doubt whether an object is
beautiful (or ugly), even when he does not doubt that some
one or other has experienced aesthetic enjoyment of (or repul-
sion to) it". And if this is not clear, this argument falls to the
(b) We may, however, fall back on Professor Moore's second
argument against this theory. If what I mean when I call an
object beautiful is that some one or other gets aesthetic enjoy-
ment from it, and if what any other person means when he calls
it ugly is that some one or other feels aesthetic repulsion to it,
we are not at variance, because both propositions might be
true. Yet if anything is clear, it is that we do suppose ourselves
to be making incompatible statements about the object.
(c) And finally, it might have been enough to ask quite
simply whether any one finds it plausible to suppose that what
I mean when I call a thing beautiful is that some one or other,
quite irrespective of who it may be, has felt aesthetic enjoyment
of it. Until some such person is found, the theory may simply
be ignored; and I have dealt with it only for the sake of logical
(2) The second theory, that what I mean is that / experience
aesthetic enjoyment from the object, is, however, often held.
The objection which Professor Moore makes to the corre-
sponding theory about Tightness is equally fatal to this theory
about beauty. If all I mean when I say that an object is beauti-
ful is that I experience aesthetic enjoyment from it, and all you
mean when you say it is ugly is that you experience aesthetic
repulsion to it, we are not at variance, for we may both be
right. Yet if anything is clear, it is that we suppose ourselves
to be making incompatible statements.
At the same time it is worth noting that, owing to the ob-
scurity of the nature of aesthetic enjoyment in comparison
with that of moral approval, each of two people who call a
thing respectively beautiful and ugly would be more inclined
to deny that the other is experiencing genuine aesthetic repulsion
or enjoyment than he would be to take the corresponding
line in an analogous moral dispute. And in many, if not in
all, cases this denial would pretty clearly be correct.
(3) The third theory is that in calling an object beautiful I
mean that a majority of some class of people experience aesthetic
enjoyment from it. To save needless repetition, I may simply
refer to the previous refutation of the corresponding theory
about goodness.1 I may add that all the historical instances of
artistic revolt against tradition, all the claims to have discovered
some new form of beauty, are clear evidence of the falsity of
the theory. Nay, any one who acclaims the beauty of a newly
discovered mountain or flower, or of a neglected gem of poetry,
is evidently not claiming to express the feelings of any majority.
Finally, theories (4) and (5), and theory (B) in all its five
possible forms, may be refuted by the same arguments as the
corresponding theories about tightness or about goodness.
We have now seen the falsity of all purely subjective theories
of beauty, i. e. of all theories which identify the beauty of an
object with its arousing of either a feeling or an opinion in any
man or class of men. But a doubt remains whether beauty can
be purely objective, whether it must not in some more com-
plicated way have necessary reference to a mind. The difficulty
arises partly from the fact that so much of beauty—the whole
of it, I suppose, strictly speaking—is sensuous. Very much of
it is inseparably bound up with colours and sounds, and unless
we give a completely objective account of these we cannot give
a completely objective account of beauty. .Now the view which
holds that colours and sounds, exactly as they are (or seem to
be) perceived, are characteristic of physical bodies, is pretty
generally given up. Those who believe in their objectivity are
now more inclined to describe them as forming a separate class
of objects to which the name ofsensa is given. And again most
« Cf. pp. 83-4,
of the sensum-theories regard sensa as owing their nature not
solely to the nature of the physical object but partly to the
nature of the perceiver's body, or of his mind, or of both. The
only form of sensum-theory which gives a purely objective
account of colours and sounds, the only one therefore that is
compatible with a purely objective account of beauty, is one
which regards them as pure products of the physical object,
and the mind's function with regard to them as being purely
that of perceiving a selection of them. This is obviously not
the place to embark upon a discussion of the "object of percep-
tion* (particularly as beauty is not one of our main subjects but
is being examined only in order to bring out by contrast the
nature of goodness). But we may perhaps be allowed to assume
that the purely selective theory is highly improbable. Person-
ally I would go further and say that to me any form of sensum-
theory seems less probable than a purely causal theory of per-
ception; one, that is to say, which regards all our sensuous
experience as not being apprehension at all but simply a set of
mental events produced by the operation of external bodies on
our bodies and through them on our minds. But the commend-
ing of this view would need a book to itself, and in anything
I say which presupposes this view I can only hope to win the
assent of those to whom such a view of perception already
commends itself.
If we are right in holding that beauty is bound up with what
is sensuous and that what is sensuous is at any rate partly de-
pendent on a mind, can any account of beauty be given which
will do justice to this fact and at the same time avoid the fatal
objections which we have seen to arise against a purely sub-
jective account of beauty, i, e. one which identifies 'this is
beautiful' with 'this arouses some particular feeling or opinion
in some mind or minds'? The view to which I find myself
driven, in the attempt to avoid the difficulties that beset both a
purely objective and a purely subjective view, is one which
identifies beauty with the power of producing a certain sort of
experience in minds, the sort of experience which we are familiar
with under such names as aesthetic enjoyment or aesthetic
thrill. Does such a theory really escape the objections to a
purely subjective theory? We shall see the position of affairs
if we suppose first two people maintaining respectively that a
certain object is beautiful and that it is not beautiful; and
secondly two people maintaining respectively that an object is
beautiful and that it is ugly, (i) When a man thinks an object
to be beautiful, either he is holding an independent personal
opinion, or he is judging the object to be beautiful out of
deference to the opinion of some other person or persons.
For brevity's sake I will omit the latter or conventional type
of opinion, which raises no special difficulty, and concentrate
on the former. Every independent opinion that an object is
beautiful has for antecedent, it will be agreed, a feeling of enjoy-
ment which either is, or is taken to be, aesthetic; there is no
other basis for thinking an object beautiful than such an ex-
perience. Yet what the judger is judging is not that he has this
eiperience. For (a) the judgement is a judgement about the
object, not about the judger's state of mind. And (^) though
we cannot judge an object to be beautiful till we think we have
been aesthetically thrilled by it, we judge that it was beautiful
before we were thrilled by it, and will be beautiful when we
have ceased to be thrilled by it. The judgement, while it is not
a judgement about the judger's state of mind, is one in which,
on the strength of his knowledge of (or opinion about)
his state of mind, he ascribes an attribute to an object. And
if we ask ourselves what is the common attribute belonging
to all beautiful objects, we can, I believe, find none other
than the power of producing the kind of enjoyment known as
Now observe the difference between this view and that
which identifies *this is beautiful* with 'this produces a certain
feeling in me*. The latter view is open to the fatal objection
that it leaves no real point at issue between the man who says
'this is beautiful* and the man who says 'no, it is not*; for it may
" I think, however, that it must be admitted (and this »Sj for what it is worth, an
objection to the view suggested) that we do not mean by 'beautiful* an attribute having
even this sort of reference to a mind, but something entirety resident in the object,
apart from relation to a'mind. What 1 am suggesting is that we are deceived in think-
ing that beautiful things have any such common attribute over and above the power of
producing aesthetic enioyiaent.
well be true both that the object produces aesthetic enjoyment
in one man and that it does not produce it in the other. But if
one is in effect saying 'the object has it in it to produce aesthetic
enjoyment in any one sufficiently capable of feeling such',-there
is a question really at issue between him and one who says 'no,
it has not*.
That this is no unnatural account of the question at issue
between them may be seen by considering separately the two
cases in which the person who asserts the object to be beautiful
is right, and is wrong, respectively, (a) When he is right, he is
actually having or has had genuine aesthetic enjoyment of the
object; the other has had no such enjoyment and is inferring
from this that the object has no power of producing such enjoy-
ment in any one. But the actual occurrence of the enjoyment
depends on conditions in the experient as well as on condi-
tions in the object. While his own lack of susceptibility
or of aesthetic education is at fault, he is supposing that the
conditions in the object are not those required for aesthetic
(If) When the man who pronounces the object beautiful is
wrong, what is happening is that he is receiving some kind of
pleasurable thrill from the object, confusing this with aesthetic
enjoyment, and on the strength of this mistake wrongly at-
tributing to the object the power of producing the aesthetic
(2) When one person pronounces an object beautiful and
another pronounces it ugly, there are, I think it must be ad-
mitted, four possibilities. The first person alone may be right,
or the second alone be right, while the other is mistaking some
other emotion for aesthetic repulsion or enjoyment, or making
a conventional judgement based on his opinion that some one
else thinks the object ugly or beautiful. Thirdly, both may be
wrong, both making the sort of mistake that in the first two
cases one of them has made. But fourthly, it appears that both
may conceivably be right. For when we consider how beauty
is bound up with sense-perception, and how dependent sense-
perception is on our equipment of sense-organs, it seems not
improbable that owing to differences in the sense-organs of
different individuals, or of different races, the same object may
produce (or minister to) genuine aesthetic enjoyment in one
individual or race and genuine aesthetic repulsion in another.
And if so, the same object will 'be both beautiful and ugly;
whether we think this probable or not, it seems clear that it
cannot be ruled out as impossible. And to this extent our
ordinary ideas about beauty and ugliness require revision; for
we certainly in general mean by 'beautiful* and "ugly* attributes
which cannot belong to the same thing (i. e, to the same thing
taken as a whole; we well know that some elements in a thing
may be beautiful and others ugly).
In a sense, then, beauty is perfectly objective. An object
may be beautiful though no one has ever felt or will ever feel
its beauty, provided that there is some mind which if con-
fronted with it would get aesthetic enjoyment from it, or could
be so educated as to get such enjoyment. The further question
may be asked, whether there could be beauty in a world in
which there were no minds at all. If it is possible for a mind to
come into being in such a world (a question on which I offer no
opinion), there too there might be beauty; for an object in a
mindless universe might be such as to produce aesthetic enjoy-
ment in some mind which might come into existence. So far
the theory is purely objective; but In so far as it holds beauty
to be indefinable except by reference to a possible effect on a
mind, and to involve a certain relation between the nature of
the object and that of mind (or of some mind), our theory
might be called subjective. And its objectivity in the one sense
seems to be quite compatible with its subjectivity in the other.
Finally we may ask whether beauty is an intrinsic value.
Apparently we must say that it is not. For even if it be held
that beauty might exist in a mindless universe, there is no
reason for regarding the beautiful things in such a universe as
having any value in themselves. Their value would be solely
instrumental to the production of aesthetic enjoyment in such
minds as might later come into being. Aesthetic enjoyment is
good in itself, and beauty is valuable simply as productive of
it; in any assignment of intrinsic value to a mindless universe
or to anything in it there is a surreptitious introduction of a

subjective factor, namely of oneself imagined as contemplating

such a universe or the things in it.

"We may turn now from beauty to goodness. We have

already seen the failure of all purely subjective theories of good-
ness. Are we, then, to adopt a purely objective theory, or an
intermediate theory such as I have suggested about beauty?
Such a theory would hold that goodness is a power in certain
things to produce a certain sort of feeling in minds—or, it
might be, to produce in minds the opinion that these things
are good. These two sub-types of theory must be considered
separately, (A) There seems to be a great difference between
our predications of beauty and of goodness. All predications
of beauty seem to rest on the opinion that the judger, or some
one else, has derived aesthetic enjoyment from the object. If
for brevity we eliminate the derivative or conventional predica-
tions of beauty in which we rely on some one else's supposed
enjoyment of the object, and if we concentrate on independent
predications of beauty, we find that it is quite impossible to
make such a predication unless we have ourselves first derived
aesthetic enjoyment from the object. The feeling is the primary
thing and the judgement is based upon it. In my attitude
towards the things I call good, on the other hand, my opinion
that they are good seems to be the primary thing, and my
feeling to be consequent upon this. It seems even evident that
I may judge something to be very good or very bad, when
owing to preoccupation I have no strong feeling about it at all;
while (apart from conventional judgements) no one can judge
an object to be very beautiful or very ugly except on the strength
of a vivid feeling towards it. This is a matter of psychological
analysis, and it is difficult to be certain about it. But after a
good deal of reflection I feel pretty clear that if I delight in
contemplating a virtuous action, for instance, it is because I
think the action to be good, and not vice versa* If we think
that feeling is here the primary thing, I suppose the most natural
name for it would be 'approval'; but if we ask ourselves what
approval is we find that the basic element in it is not feeling at
all but the judgement that an object is good.
Further, the chaotic condition of aesthetic theory seems to
show that it is extremely difficult, if not impossible, to specify
the characteristic or characteristics of beautiful things (apart
from the way in which they can affect minds) on which beauty
depends. But we can give an account of what the goodness of
good things depends upon without introducing any reference
to the feelings they produce in a spectator. The goodness of
an action depends on its springing from one or other of certain
nameable motives; the goodness of an intellectual activity,
on its being an instance of knowledge, or of well-grounded
opinion; the goodness of a state of feeling, on its being pleasant,
(ff) If goodness is not a power of producing a certain feeling,
is it a power of producing a certain opinion—the opinion that
the object is good ? This suggestion can be refuted much more
decisively. For opinions have the characteristic, which feelings
have not, of being necessarily either true or false. No one will
say that an object is good because it is capable of so affecting
him, or any one, as to produce the opinion that it is good,
irrespective of whether the opinion is true or false. If the
opinion is false, the object is not good; and if the opinion is
true, the truth of the opinion that the object is good rests on
the/act that the object is good, and not vice versa.
It seems then that an intermediate view of goodness is no
more satisfactory than a purely subjective view, and that good-
ness is entirely objective and intrinsic to the things that are
What are we to say of Tightness? We must, I think, say
that it is intrinsic, but that in so far as a right act has value its
value is not intrinsic. The Tightness of an act, if the contention
of our second chapter is correct, is intrinsic to the act, depending
solely on its nature. But if we contemplate a right act alone,
it is seen to have no intrinsic value. Suppose for instance that it
is right for a man to pay a certain debt, and he pays it. This is
in itself no addition to the sum of values in the universe. If he
does it from a good motive, that adds to the sum of values in
the universe; if from a morally indifferent motive, that leaves
the sum of values unchanged; if from a bad motive, that de-
tracts from the balance of values in the universe. Whatever
intrinsic value, positive or negative, the action may have, it
owes to the nature of its motive and not to the act's being right
or wrong; and whatever" value it has independently of its motive
is instrumental value, i.e. not goodness at all, but the property
of producing something that is good.

Our arguments seem to show, then, that Tightness no more

than beauty is an intrinsic value, and that the only correct way
of expressing intrinsic value is the word 'good*.

O UR next step is to inquire what kinds of thing are intrinsi-

cally good, (i) The first thing for which 1 would claim
that it is intrinsically good is virtuous disposition and action,
i, e, action, or disposition to act, from any one of certain motives,
of which at all events the most notable are the desire to do one's
duty, the desire to bring into being something that is good,
and the desire to give pleasure or save pain to others. It seems
clear that we regard all such actions and dispositions as having
value in themselves apart from any consequence. And if any one
is inclined to doubt this and to think that, say, pleasure alone is
intrinsically good, it seems to me enough to ask the question
whether, of two states of the universe holding equal amounts of
pleasure, we should really think no better of one in which the
actions and dispositions of all the persons in it were thoroughly
virtuous than of one in which they were highly vicious. To
this there can be only one answer. Most hedonists would shrink
from giving the plainly false answer which their theory requires,
and would take refuge in saying that the question rests on a
false abstraction. Since virtue, as they conceive it, is a dis-
position to do just the acts which will produce most pleasure,
a universe full of virtuous persons would be bound, they might
say, to contain more pleasure than a universe full of vicious
persons. To this two answers may be made, (a) Much pleasure,
and mudh pain, do not spring from virtuous or vicious actions
at all but from the operation of natural laws. Thus even if a
universe filled with virtuous persons were bound to contain
more of the pleasure and less of the pain that springs from
human action than a universe filled with vicious persons would,
that inequality of pleasantness might easily be supposed to be
precisely counteracted by, for instance, a much greater inci-
dence of disease. The two states of affairs would then, on
balance, be equally pleasant; would they be equally good?
And (l>) even if we could not imagine any circumstances in
which two states of the universe equal in pleasantness but
unequal In virtue could exist, the supposition is a legitimate
one, since it is only intended to bring before us in a vivid way
what is really self-evident, that virtue is good apart from its
(2) It seems at first sight equally clear that pleasure is good
in itself. * Some will perhaps be helped to realize this if they
make the corresponding supposition to that we have just made;
if they suppose two states of the universe including equal
amounts of virtue but the one including also widespread and
intense pleasure and the other widespread and intense pain.
Here too it might be objected that the supposition is an im-
possible one, since virtue always tends to promote general
pleasure, and vice to promote general misery. But this objection
may be answered just as we have answered the corresponding
objection above.
Apart from this, however, there are two ways in which even
the most austere moralists and die most anti-hedonistic philo-
sophers are apt to betray the conviction that pleasure is good in
itself, (a) One is the attitude which they, like all other normal
human beings, take towards kindness and towards cruelty. If
the desire to give pleasure to others is approved, and the desire
to inflict pain on others condemned, this seems to imply the
conviction that pleasure is good and pain bad. Some may
think, no doubt, that the mere thought that a certain state of
affairs would be painful for another person is enough to account
for our conviction that the desire to produce it is bad. But I
am inclined to think that there is involved the further thought
that a state of affairs in virtue of being painful is prima facie
(i. e. where other considerations do not enter into the case) one
that a rational spectator would not approve, L e. is i>ad; and
that similarly our attitude towards kindness involves the
thought that pleasure is good. (£) The other is the insistence,
which we find in the most austere moralists as in other
people, on the conception of merit. If virtue deserves to be
rewarded by happiness (whether or not vice also deserves to
be rewarded by unhappiness), this seems at first sight to imply
that happiness and unhappiness are not in themselves things
indifferent, but are good and bad respectively.
Kant's view on this question is not as clear as might be
wished. He points out that the Latin konum covers two notions,
distinguished in German as das Gute (the good) and Jos VFohl
(well-being, i. e, pleasure or happiness); and he speaks of'good'
as being properly applied only to actions,1 i,e, he treats 'good*
as equivalent to 'morally good', and by implication denies that
pleasure (even deserved pleasure) is good. It might seem then
that when he speaks of the union of virtue with the happiness
it deserves as the bonum consummatum he is not thinking of
deserved happiness as good but only as das Wokl, a source
of satisfaction to the person who has it. But if this exhausted
his meaning, he would have no right to speak of virtue, as he
repeatedly does, as das oberste Gut; he should call it simply das
Gute, and happiness das Wold, Further, he describes the union
of virtue with happiness not merely as 'the object of the desires
of rational finite beings', but adds that it approves itself 'even
in the judgement of an impartial reason* as "the whole and
perfect good', rather than virtue alone. And he adds that
'happiness, while it is pleasant to the possessor of it, is not of
itself absolutely and in all respects good, but always presupposes
morally right behaviour as its condition'; which implies that
when that condition is fulfilled, happiness is good.2 All this
seems to point to the conclusion that in the end he had to recog-
nize that while virtue alone is morally good, deserved happiness
also is not merely a source of satisfaction to its possessor, but
objectively good.
But reflection on the conception of merit does not support the
view that pleasure is always good in itself and pain always bad
in itself. For while this conception implies the conviction that
pleasure when deserved is good, and pain when undeserved
bad, it also suggests strongly that pleasure when undeserved is
bad and pain when deserved good.
There is also another set of facts which casts doubt on the
view that pleasure is always good and pain always bad. We
have a decided conviction that there are bad pleasures and
Kritifc Jerpr, ftntanftt 59-60 (Akad. Ausgabe, vol. v), 150—1 (Abbott's Trans.,
ed,1 6).
Ib. no-ii (Akad, Ausgabe), 206-7 (Abbott),
(though this is less obvious) that there are good pains. We
think that the pleasure taken either by the agent or by a spectator
in, for instance, a lustful or cruel action is bad; and we think it
a good thing that people should be pained rather than pleased
by contemplating vice or misery.
Thus the view that pleasure is always good and pain always
bad, while it seems to be strongly supported by some of our
convictions, seems to be equally strongly opposed by others.
The difficulty can, I think, be removed by ceasing to speak
simply of pleasure and pain as good or bad, and by asking more
carefully what it is that we mean. Consideration of the question
is aided if we adopt the view (tentatively adopted already)1
that what is good or bad is always something properly ex-
pressed by a that-clause, i. e. an objective, or as I should prefer
to call it, A fact. If we look at the matter thus, I think we can
agree that the fact that a sentient being is in a state of pleasure
is always in itself good, and the fact that a sentient being is in a
state of pain always in itself bad, when this fact is not an element
in a more complex fact having some other characteristic relevant
to goodness or badness. And where considerations of desert
or of moral good or evil do not enter, i. e, in the case of animals,
the fact that a sentient being is feeling pleasure or pain is the
whole fact (or the fact sufficiently described to enable us to
judge of its goodness or badness), and we need not hesitate to
say that the pleasure of animals is always good, and the pain of
animals always bad, in itself and apart from its consequences.
But when a moral being is feeling a pleasure or pain that is
deserved or undeserved, or a pleasure or pain that implies a
good or a bad disposition, the total fact is quite inadequately
described if we say 'a sentient being is feeling pleasure, or pain*.
The total fact may be that 'a sentient and moral being is feel-
ing a pleasure that is undeserved, or that is the realization of a
vicious disposition', and though the fact included in this, that
*a sentient being is feeling pleasure' would be good if it stood
alone, that creates only a presumption that the total fact is good,
and a presumption that is outweighed by the other element in
the total fact.
> pp. 111-13.
Pleasure seems, indeed, to have a property analogous to
that which we have previously recognized under the name of
conditional orprima facie tightness. An act of promise-keeping
has the property, not necessarily of being right but of being
something that is right if the act has no other morally significant
characteristic (such as that of causing much pain to another
person). And similarly a state of pleasure has the property,
not necessarily of being good, but of being something that is
good if the state has no other characteristic that prevents it
from being good. The two characteristics that may interfere
with its being good are (a) that of being contrary to desert, and
(S) that of being a state which is the realization of a bad dis-
position. Thus the pleasures of which we can say without
doubt that they are good are (i) the pleasures of non-moral
beings (animals), (ii) the pleasures of moral beings that are
deserved and are either realizations of good moral dispositions
or realizations of neutral capacities (such as the pleasures of
the senses).
In so far as the goodness or badness of a particular pleasure
depends on its being the realization of a virtuous or vicious
disposition, this has been allowed for by our recognition of
virtue as a thing good in itself. But the mere recognition of
virtue as a thing good in itself, and of pleasure as a thing prima
facie good in itself, does not do justice to the conception of
merit. If we compare two imaginary states of the universe,
alike in the total amounts of virtue and vice and of pleasure and
pain present in the two, but in one of which the virtuous were
all happy and the vicious miserable, while in the other the
virtuous were miserable and the vicious happy, very few people
would hesitate to say that the first was a much better state of
the universe than the second. It would seem then that, besides
virtue and pleasure, we must recognize (3), as a third inde-
pendent good, the apportionment of pleasure and pain to the
virtuous and the vicious respectively. And it is on the recog-
nition of this as a separate good that the recognition of the duty
of justice, in distinction from fidelity to promises on the one
hand and from beneficence on the other, rests.
(4) It seems clear that knowledge, and in a less degree what
we may for the present call 'right opinion", are states of mind
good in themselves. Here too we may, if we please, help our-
selves to realize the fact by supposing two states of the universe
equal in respect of virtue and of pleasure and of the allocation
of pleasure to the virtuous, but such that the persons in the one
had a far greater understanding of the nature and laws of the
universe than those in the other. Can any one doubt that the
first would be a better state of the universe ?
From one point of view it seems doubtful whether knowledge
and right opinion, no matter what it is of or about, should be
considered good. Knowledge of mere matters of fact (say of
the number of stories in a building), without knowledge of
their relation to other facts, might seem to be worthless; it
certainly seems to be worth much less than the knowledge of
general principles, or of facts as depending on general principles
—what we might call insight or understanding as opposed to
mere knowledge. But on reflection it seems clear that even
about matters of fact right opinion is in itself a better state of
mind to be in than wrong, and knowledge than right opinion.
There is another objection which may naturally be made to
the view that knowledge is as such good. There are many pieces
of knowledge which we in fact think it well for people not to
have; e. g. we may think it a bad thing for a sick man to know
how ill he is, or for a vicious man to know how he may most
conveniently indulge his vicious tendencies. But it seems that
in such cases it is not the knowledge but the consequences in
the way of pain or of vicious action that we think bad.
It might perhaps be objected that knowledge is not a better
state than right opinion, but merely a source of greater satis-
faction to its possessor. It no doubt is a source of greater
satisfaction. Curiosity is the desire to know^ and is never really
satisfied by mere opinion. Yet there are two facts which seem
to show that this is not the whole truth, (a) While opinion
recognized to be such is never thoroughly satisfactory to its
possessor, there is another state of mind which is not knowledge
—which may even be mistaken—yet which through lack of
reflection is not distinguished from knowledge by its possessor,
the state of mind which Professor Cook Wilson has called
'that of being under the impression that so-and-so is the case*.1
Such a state of mind may be as great a source of satisfaction to
its possessor as knowledge, yet we should all think it to be an
inferior state of mind to knowledge. This surely points to a
recognition by us that knowledge has a worth other than that
of being a source of satisfaction to its possessor. ($) Wrong
opinion, so long as its wrongness is not discovered, may be as
great a source of satisfaction as right. Yet we should agree
that it is an inferior state of mind, because it is to a less extent
founded on knowledge and is itself a less close approximation
to knowledge; which again seems to point to our recognizing
knowledge as something good in itself.
Four things, then, seem to be intrinsically good—virtue,
pleasure, the allocation of pleasure to the virtuous, and know-
ledge (and in a less degree right opinion). And I am unable to
discover anything that is intrinsically good, which is not either
one of these or a combination of two or more of them. And
while this list of goods has been arrived at on its own merits,
by reflection on what we really think to be good, it perhaps
derives some support from the fact that it harmonizes with a
widely accepted classification of the elements in the life of the
soul. It is usual to enumerate these as cognition, feeling, and
conation. Now knowledge is the ideal state of the mind, and
right opinion an approximation to the ideal, on the cognitive
or intellectual side; pleasure is its ideal state on the side of feel-
ing; and virtue is its ideal state on the side of conation; while
the allocation of happiness to virtue is a good which we recog-
nize when we reflect on the ideal relation between the conative
side and the side of feeling. It might of course be objected
that there are or may be intrinsic goods that are not states of
mind or relations between states of mind at all, but in this
suggestion I can find no plausibility. Contemplate any imagi-
nary universe from which you suppose mind entirely absent,
and you will fail to find anything in it that you can call good in
itself. That is not to say, of course, that the existence of a
material universe may not be a necessary condition for the
existence of many things that are good in themselves. Our
' Statement aaj Inference, i 113,
knowledge and our true opinions are to a large extent about
the material world, and to that extent could not exist unless it
existed, Our pleasures are to a large extent derived from
material objects. Virtue owes many of its opportunities to the
existence of material conditions of good and material hindrances
to good. But the value of material things appears to be purely
instrumental, not intrinsic.
Of the three elements virtue, knowledge, and pleasure are
compounded all the complex states of mind that we think good
in themselves. Aesthetic enjoyment, for example, seems to be
a blend of pleasure with insight into the nature of the object
that inspires it. Mutual love seems to be a blend of virtuous
disposition of two minds towards each other, with the know-
ledge which each has of the character and disposition of the
other, and with the pleasure which arises from such disposition
and knowledge. And a similar analysis may probably be applied
to all other complex goods.

T HE next question which seems to face us is the question

whether goods are commensurable. As a preliminary to
this, we may ask the narrower question whether pleasures are
commensurable. And as a preliminary to that again, we may
ask whether pleasures are comparable; whether we can properly
say that one pleasure is greater or more pleasant than another.
That this is so there seems to be no doubt. Any denial of it
usually proceeds by asking us to say which of two extremely
disparate pleasures is the greater. It is true that often we cannot
do this. But one who maintains the comparability of pleasures
is not bound to maintain that he can say of any two pleasures
which is the greater. He may not, for instance, have recently
enjoyed the two; and undoubtedly memory of the intensity of
pleasures, is treacherous when it has to cover any long span of
rime. Or again one or both of the pleasures may be very com-
plex, and the introspection necessary for the assessment of
their pleasantness may be difficult. But this much is clear, that
if we compare two comparatively simple and comparatively
similar pleasures recently experienced, we may have no difficulty
in seeing one to have been more pleasant than the other; and
this is enough to show that pleasures have different intensities.
If we proceed from these to others that are more complex or
less similar or were experienced less recently, it is plain that
there is no point at which we can stop and say, some pleasures
are in their nature comparable and others are not. The truth
is that pleasures have the only characteristic on their side that
is necessary to comparability, viz. difference of intensity; but
that some of the conditions on our side which would enable
us to compare them are often lacking.1
If this is clear, we may next argue that pleasures are com-
» It Is hardly necessary to say that I am not ascribing to pleasures an existence inde-
pendent of a mind to fee! them; I am thinking of them as existing independently of
our activity in comparing them.
mensurable,!, e. that one pleasure may not only be more intense
than another, but may be, say, just twice as intense. This
surely follows from our previous conclusion. For if one
pleasure is more intense than another, it must have a definite
amount of extra intensity. True (as is often pointed out), the
difference between two pleasures is not a pleasure, as the differ-
ence between two lengths is a length. That is part of what we
mean by distinguishing intensity from extension. But the
difference between the intensity of two pleasures is an amount
of intensity, and of course a definite amount. And this may be
exactly equal to the intensity of the less pleasure; and if so, the
greater pleasure is just twice as great as the less.
On their side, then, pleasures have the characteristic requisite
for coraniensurabiltty. And on our side too the conditions
necessary for rough commensuration are sometimes fulfilled.
For there seenis to be no doubt that in comparing two com-
paratively simple and similar pleasures, recently experienced, we
can sometimes say with the greatest confidence that one was, say,
at least twice as intense as the other. We cannot be more precise
than this and say that one was just twice, or three times, as
intense as the other. That points to the absence of some of the
conditions needed for exact commensuration on our side and
not to the lack of the conditions needed for commensurability
on theirs. Or perhaps this statement should be modified. The
pleasures have precise intensities, which in principle makes
them commensurable. But they have not the characteristics
that make material objects comparatively easy to measure, the
characteristics of being fairly permanent and of being easily
compared with standards of measurement which may be applied
to each in turn.
At the same time it is worth while to point out that the
difference between physical and mental objects in respect of
commensurability is not so complete as it might appear. We
can never say that one material object is precisely twice as long,
say, as another, but only at most that it either is this or fails
to be this by an amount which our senses do not enable us to
detect. And similarly we can, under favourable circumstances,
say that one pleasure either is just twice as intense as another,
or differs from this by an amount which introspection does
not enable us to detect.
If pleasures are commensurable, it does not necessarily follow
that all goods are; that, say, pleasure can be measured in respect
of goodness against virtue or knowledge. But the arguments
that have been urged against the eornmensurability of goods
are to a large extent the same that have been urged against the
eommensurabiiity of pleasures. The case against the eom-
mensurabiiity of goods gets much of its plausibility by con-
fronting us with two goods, highly complex or highly dissimilar
or not recently experienced, or perhaps never experienced at all
by the person who is asked to give judgement, or a good which
ranks high in a low order with one which ranks low in a higher
order,1 The proper reply is to point to other goods not so
characterized which we have little difficulty in comparing and
even in measuring roughly against one another. Few people,
except those who are committed to hedonism, would hesitate
to say that an action of sublime heroism is many times as good
a thing as some faint and passing shade of pleasure. If we can
ever commeasure two goods, that is much stronger evidence
that goods are in principle commensurable, than the fact that
we sometimes cannot commeasure two goods is for goods not
being in principle commensurable.2 Yet the conclusion to
which this preliminary consideration points is faced with great
difficulties when we try to commeasure good things of very
different types.
When we turn to the attempt actually to measure goods
against each other, it will be prudent to begin with goods of
the same order, i. e, with the balancing of one pleasure against
another, of one state of knowledge against another, of one
virtuous action or disposition against another, (i) Much has
been written about the calculus of pleasures, and I shall be
forgiven, and perhaps thanked, for not saying much more.
Bentham has enumerated the 'dimensions' on which depend
the calculus of pleasures, but it is evident that two of them—
« Cf.pp, 149-54.
I note with satisfaction that Mr, McTaggart argues in favour of the cointnensina-
MJity of pleasures, and of goods (Tie Naturt {/Existence, ii. 14*-], 447-8),
fecundity and purity—refer not to the intrinsic value of pleasures
but to their tendency to produce other pleasures, or pains; that
certainty and propinquity, again, do not belong to pleasures in
themselves; and that extension refers not to the intrinsic value of a
particular pleasure but to the number of peoplefor whom pleasure
will be produced by a particular act. We are left with intensity
and duration as the characteristics on which depend the value of
a pleasure qua pleasure; and to these there is nothing to add.
Mill argued that pleasures have "quality" as well, and that this
affects their hedonistic value. The first of these contentions is
true, and the second false. It is obviously true that a pleasure,
i e. a pleasant state of mind, always has a quality over and above
its intensity and duration. If we compare, for instance, the
pleasure of reading good literature and that of a country walk
(even supposing that in some case their intensity and duration
were the same), we do not find a pleasant feeling identical in
the two plus cognitive and conational elements different in the
two. Pleasant feeling is coloured through and through by what
we are taking pleasure in. But Mill should have gone on to say
that however interesting a fact this may be, it is from the point
of view of the hedonistic calculus quite irrelevant. For a
hedonist holds that pleasantness is the only characteristic of
states of mind that makes them good, and one state of rnind
can have more pleasantness than another only by being more
intensely pleasant or by being pleasant for a longer time. Mill's
introduction of quality of pleasures into the hedonistic calculus
is, as has been widely recognized, an unconscious departure
from hedonism and a half-hearted admission that there are
other qualities than pleasantness in virtue of which states of
mind are good.
(2) The commensuration of different states of knowledge
against each other, and of states of knowledge against states of
right opinion, has been much less discussed. It is clear that we
all do in practice treat some forms of knowledge as more valu-
able than others, but we do not as a rule stop to distinguish
whether we are valuing them for their own nature or for their
results. Our concern here is solely with their intrinsic value.
The problem is complicated by the fact that while we assign
value, and a kindred (though not equal) value, both to know-
ledge and to right opinion, these are not species of a single
genus, Knowledge is apprehension of fact, and right opinion
is not that, but is simply a state of mind in which things are
believed (not apprehended) to be related as they are in fact
related; and there seems to be no single genus of which these
are species. We cannot say that knowledge and right opinion
are species of a single genus (say, 'cognition'), differing in degree
of certainty; for knowledge alone has certainty, and opinion
has merely varying degrees of approach to certainty. What
we can say, by way of bringing knowledge and opinion into
relation with one another, is (a) that knowledge is the ideal we
always have in mind in the forming of opinions, and opinions
are the results of more or less unsuccessful attempts at reaching
knowledge, and (£) that opinion always rests in part on know-
ledge. These two characteristics are closely bound up together,
for while opinions are the result of a variety of psychological
causes (largely irrational), it is just in so far as they rest on
knowledge that they approximate to being knowledge.
It seems natural to say that the value of a state of knowledge
or opinion rests partly on the nature of the state of mind, and
partly on the nature of the fact it is in relation to. We must not
say 'of the fact apprehended*, for in opinion the fact is not
apprehended. If we want to substitute something more definite
for the phrase 'the fact it is in relation to*, we can only say 'the
fact apprehended, or believed to exist'.
Now (a) in respect of the nature of the state of mind, there
appear to be two characteristics which mark knowledge off
from opinion, and on which the superiority of knowledge
seems to rest, (i) The first is that knowledge is either the direct
non-inferential apprehension of fact, or the inferential appre-
hension of one fact as necessitated by other facts, while opinion
is never either completely non-inferential or completely inferen-
tial, but is always the holding of a view which is partly
grounded on apprehension of fact and partly the product of
other psychical events such as wishes, hopes, fears, or the
mere association of ideas. In this respect all knowledge is
of equal value, and opinion is of value as it approximates to
being completely grounded on apprehension of fact, i, e. to
being knowledge, (ii) The other respect in which knowledge
differs from and is superior to opinion is that of certainty or
complete absence of doubt. But we cannot say that opinion
approximates in value to knowledge in proportion as it ap-
proaches to certainty. It rises in value not by being held with
more conviction, but by being held with a degree of conviction
which answers more closely to the degree to which the opinion
is grounded on knowledge. It might seem as if it were a worse
intellectual fault to hold an opinion with more, than to hold it
with less, conviction than it deserves. It is certainly a fault to
which we are more liable, and it is probably more dangerous
in its consequences; but it does not seem to be either a greater
or a less fault. To mis-estimate the probability of our opinions
is equally bad, whether we over-estimate it or under-estimate it.
Thus in what renders a state of opinion good there are two
elements, (i) its degree of groundedness, and (ii) the extent to
which the degree of conviction corresponds to the degree of
It may be added that knowledge can only be possessed with
the degree of conviction it deserves, i. e. complete conviction.
An opinion on the other hand can rarely, if ever, be held with
the precisely appropriate degree of conviction; to make sure
of that we should have to know factors in the situation which
ex hypothesi we do not know. Thus in respect (ii) as well as
in respect (i) opinion is in principle inferior to knowledge, and
can only by accident fail to be so,
But ($) our states of knowledge and opinion seem to derive
some of their value from the nature of the fact apprehended,
or believed to exist. The only rule I have to suggest here is
that which has been stated impressively by Mr, Bradley,1
that knowledge of general principles is intellectually more
valuable than knowledge of isolated matters of fact, and that
the more general the principle—the mure facts it is capable of
explaining—the better the knowledge. Our ideal in the pursuit
of knowledge is system, and system involves the tracing of
consequents to their ultimate grounds. Our aim is to know
Principles of Logic, ed. z, ii, 68J-8,
not only the 'that* but the 'why* also, when the 'that* has a
'why*. I venture, however, to think that it Is a mistake to rest
the value of knowledge on its being inferential. Inferential
knowledge, and its value, are entirely inseparable from non-
inferential knowledge, first of the ultimate premisses from
which inference must start, and secondly of the fact that the
premisses warrant the conclusions. It is better to know why
a thing is so than merely that it is so, not because your know-
ledge is inferential, but because you are knowing more, and
knowing what is more general. Any attempt to elevate either
inferential knowledge above non-inferential, or non-inferential
above inferential, is fundamentally mistaken.
There are no doubt particular matters of fact which it is
more important for our welfare that we should know, than
many general principles. It is very likely more important for
our happiness to know the characters of the people we live
among (if we can be said ever to know them) than to know
highly general mathematical principles. But this is an instru-
mental value. From the purely intellectual point of view, with
which alone we are concerned in speaking of the intrinsic value
of knowledge, generality (I venture to think) is the only valu-
able element in knowledge or opinion arising from the nature
of the fact apprehended or believed in,
The intrinsic value of a state of knowledge or opinion seems,
then, to depend on three elements, (i) the degree of its groun-
dedness on fact, (ii) the degree to which the strength of con-
viction with which it is held corresponds to its groundedness,
(iii) the generality of the fact known, or believed to exist. How
these factors are to be balanced against one another in estimat-
ing the value of any knowledge or opinion I do not profess to
know. The question is practically less important than it might
seem to be. In choosing a subject of study, we may choose
either one (like metaphysics) in which we are unlikely to obtain
much knowledge or even fairly grounded opinion, but in which
the knowledge or opinion if obtained would be more general,
or one (such as chemistry) in which we are more likely to gain
knowledge or fairly grounded opinion, but about something
less general. But what we do in such a case is not to try to
estimate the value of what we are likely to gain, but to follow
our strongest interests; and in this we are justified, for interest
is the most important condition of our gaining anything valu-
able at all
(3) As regards the balancing of virtuous actions and dis-
positions against one another, I would refer to my next chapter,
which contains all I have to say on the subject.
The 'greatest wave* now awaits us—the question whether
virtue, knowledge, and pleasure can be compared with one
another in value, and whether they can be measured against
one another. I do not pretend that the views I shall express are
certainly true, still less that I can prove them to be so. I will
only say that they are the result of a good deal of reflection
about the comparative value of these things, and that they agree,
so far as I can judge, with the views of many others who have
reflected on it. I think, then, that pleasure is definitely inferior
in value to virtue and knowledge. There are certain facts which
support this view, (i) Most people are convinced that human
life is in itself something more valuable than animal life, though
it seems highly probable that the lives of many animals contain
a greater balance of pleasure over pain than the lives of many
human beings. Most people would accept Mill's dictum that
'it is better to be a human being dissatisfied than a pig satisfied",1
though they may think such a view inconsistent with Mill's
own principles. The very fact that he felt bound to make an
admission so fatal to his cherished hedonism is some testimony
to the truth of the admission. (2) Many people whose opinion
deserves the greatest respect have undoubtedly thought that
the promotion of the general happiness was the highest possible
ideal. But the happy state of the human race which they aimed
at producing was such a state as the progress of civilization
naturally leads us to look forward to, a state much of whose
pleasantness would spring from such things as the practice of
virtue, the knowledge of truth, and the appreciation of beauty.
The ideal has owed its attractiveness in a large degree not to its
being a state of maximum happiness, but to its being a state
whose happiness would spring from such sources; and if they
• UtiL, 14 (copyright editions).
thought that the state of maximum happiness would be one
whose happiness sprang from such things as the indulgence of
cruelty, the light-hearted adoption of ill-grounded opinions,
and enjoyment of the ugly, they would immediately reject such
an ideal. They would no doubt argue that such a state as I have
just described could not in fact be one of maximum happiness.
But that, if true, is simply a consequence of the laws of the
world we live in, and does not absolve them from facing the
problem, what if the laws of nature were such as to make such
a life the most pleasant possible ? Would they then prefer it to
a state less pleasant but more virtuous and intelligent ? They
have, in fact, in pronouncing pleasure to be the sole good, not
had in mind mere pleasant consciousness, such as we suppose
animals to enjoy, but to a very large extent what I have
called1 good (i. e. morally good) pleasures, those which are
themselves actualizations of morally good dispositions (such
as love for other men and love of truth and beauty) and
which owe their goodness much more to this than to their
This argument, if it is sound, only proves that a certain
amount of virtue and knowledge would more than outweigh
a certain amount of pleasure, and it might still be held that a
certain larger amount of pleasure would more than outweigh
the given amount of virtue and intelligence. But if we take
this view we are faced by the question, what amount of pleasure
is precisely equal in value to a given amount of virtue or of
knowledge? And to this question, so long as we think that
some amount is equal, 1 see no possibility of an answer or of an
approach to one. With regard to pleasure and virtue, it seems
to me much more likely to be the truth that no amount of
pleasure is equal to any amount of virtue, that in fact virtue
belongs to a higher order of value, beginning at a point higher
on the scale of value than that which pleasure ever reaches |
in other words, that while pleasure is comparable in value with
virtue (i. e. can be said to be less valuable than virtue) it is not
commensurable with it, as a finite duration is not commensurable
with infinite duration.* Pleasure will then bean object worthy
' Cf. p. 166,
of production, but only when this does not interfere with the
production of virtue.
(3) The question seems to become clearer when one turns
from considering virtue and pleasure in the abstract to consider
which of them seems the most worth while to get for ourselves.
It seems clear that, viewed in this way, pleasure reveals itself as
a cheap and ignoble object in comparison with virtue. This
manifests itself clearly in the fact that the acquisition of pleasure
for oneself rarely, if ever, presents itself as a duty, and usually
only as something permissible when it does not interfere with
duty, while the attainment of moral goodness habituallypresents
itself as a duty. This surely points to an infinite superiority of
virtue over pleasure, a superiority such that no gain in pleasure
can make up for a loss in virtue.
But if virtue is better worth aiming at for ourselves than
pleasure, it is better worth trying to promote for man in general.
For that which is good owes its goodness not to being possessed
by one person or by another, but to its nature.
(4) A further argument in favour of this view seems to arise
from the consideration of pleasures which are realizations of a
bad disposition, such as the pleasure of cruelty. It seems clear
that when we consider such a pleasure we are able to say at once
that it is bad, that it would have been better that it should not
have existed. If the goodness of pleasure were commensurable
with the goodness or badness of moral disposition, it would
be possible that such a pleasure if sufficiently intense should be
good on the whole. But in fact its intensity is a measure of its
badness, because it is a measure of the viciousness of the dis-
position realized.
(5) It is, on the face of it, much less obvious that knowledge
completely transcends pleasure in value than that virtue does
so. In fact it would be paradoxical to say that the slightest
possible increase in knowledge would outweigh the greatest
possible loss of pleasure; I do not think that this can reasonably
be maintained about knowledge considered simply as a con-
dition of the intellect. But in fact most (if not all) states of know-
ledge are themselves to some extent the actualization of a desire
for knowledge. And the desire or disposition expressed and
manifested in them has moral worth, is of the nature of virtue,
and has thus a value completely transcending that of pleasure
(i, e. pleasure that is the actualization not of a good disposition
but of a neutral capacity, as sensuous pleasure is). It is to be
noted, however, that this desire is expressed just as much in the
search for knowledge as in the attainment of it. While in its
own nature knowledge seems to be a better state than inquiry,
the moral virtue of desire for knowledge may be much more
fully present in many an unsuccessful search for difficult
knowledge than in the successful attainment of knowledge
that is easy to get.
The view that virtue and knowledge are much better things
than pleasure does not in practice work out in so ascetic a
way as might appear, and that for two reasons, (r) It is quite
certain that by promoting virtue and knowledge in ourselves
and in others we shall inevitably produce much pleasant con-
sciousness. These are, by general agreement, among the surest
sources of happiness for their possessors. And still more are
they among the surest sources of general happiness, for a
variety of reasons which will readily suggest themselves.
And (2) it is pretty clear that our pursuit of the greater goods
is made all the more effective by intervals in which we give
ourselves up to enjoying ourselves and helping others to enjoy
themselves. The desire for pleasure is so strong in every one
that any one who tries to ignore or suppress it entirely
will find himself defeated by the laws of human nature;
tuauram expellas furca, tumen usque recurret. There is a place
for asceticism in the best life, but it cannot safely be made the
general rule of the best life.
When we turn to consider the relative value of moral good-
ness and knowledge as ends, here again I am inclined to think
that moral goodness is infinitely better than knowledge. Here
too the question seems to become clearer when one considers
these two goods not in the abstract but as objects to be striven
after for ourselves as individuals. When I ask myself whether
any increase of knowledge, however great, is worth having at
the cost of a wilful failure to do my duty or of a deteriora-
tion of character, I can only answer in the negative. The infinite
superiority of moral goodness to anything else is clearest in
the case of the highest form of moral goodness, the desire to do
one's duty. But even of the lesser virtues the same appears to
be true. And if virtue is always the thing best worth aiming at
for oneself, it is the thing best worth trying to promote in
This doctrine will no doubt appear to many unduly mora-
listic; and two remarks may be made which may perhaps do
something to diminish their repugnance to it. (i) I do not wish
to introduce moralistic considerations into other spheres than
that of morals. Scientific questions must be dealt with on
scientific grounds, and aesthetic questions on aesthetic grounds.
That a belief would conduce to the promotion of morality does
not tend to show that it is true; and that the contemplation of
a work of art would conduce to immorality does not show that
it is not beautiful. But science and aesthetic enjoyment fall
within human life as a whole, and I do maintain that in human
life there is a greater good than either.
(2) The doctrine that morality is entirely social, that all duty
consists in promoting the good of others, seems to me a pro-
found mistake. Intellectual integrity, the love of truth for its
own sake, is among the most salient elements in a good moral
character. It is a thing which, with the other virtues, we should
try to cultivate in ourselves and to promote in others; and it
will perhaps not seem paradoxical to say that it is a better thing
to love truth than it would be to have it (if we could do so)
without loving it. That which (I am suggesting) is less good
than virtue is not the intellectual life in its concreteness, which
is the manifestation of a high and a precious excellence of
character, but the bare being in possession of knowledge irre-
spective of the character front which this springs.
The superiority of virtue to all other goods is illustrated
once more if we compare it with the more complex good we
have recognized, consisting of the proportionment of pleasure
to virtue. For suppose that some one's life as a whole contained
just that amount of pleasure which was appropriate to the
degree of virtue that characterized him, but that some of his
pleasures were the expressions of a bad disposition (e. g. pleasures
of cruelty), we should, I think, certainly judge that the state of
affairs would have been better If he had been without both
these pleasures and the defect of character they presuppose;
although the state of affairs thus produced would have been
one of much less perfect proportionment of pleasure to virtue,
since he would both have been more virtuous and have enjoyed
less pleasure.
The suggestion that there are two orders or classes of good
things such that those in one class are not commensurable,
though they are comparable, with those in the other, is obviously
not free from difficulties. But it is the conclusion to which we
are led if it be admitted that on the one hand virtue and, say,
pleasure are both of them good, and that on the other we are
totally unable to see how any amount of one of these could be
equal in goodness to any amount of the other. The problem
seems to me not to have been faced by those who repudiate
both the Kantian view that only virtue is good in itself and the
hedonistic view that only pleasure is good in itself; and I am
satisfied to have called attention to the importance and difficulty
of the problem. Its theoretical importance is evident; but it is
also practically important, for pretty evidently one of our chief
duties is that of producing as much good as we can, and clear
views as to the comparative goodness of different kinds of good
things are an essential preliminary to this.

G is good is good by virtue of something else in its nature,


by being of a certain kind, I think that I can best explain what

I understand by 'morally good' by saying that it means 'good
either by being a certain sort of character or by being related
in one of certain definite ways to a certain sort of character'.
It seems necessary to put the matter in this alternative way,
because a variety of kinds of thing can be said to be morally
good. We may say that such and such a man is morally good
or that a certain action or a certain sort of feeling (e.g. sympathy
with misfortune) is morally good. But it seems clear that a
man is morally good by virtue of having a character of a certain
kind, and that an action or a feeling is morally good by virtue
of proceeding from a character of a certain kind. This account
of the meaning of 'morally good' seems to be confirmed by
studying the usage of the phrase. Suppose it were said, for
instance, that while conscientious action, knowledge, and
pleasure are all good, conscientious action is the only one of
the three that is morally good, what we should mean is that
while conscientious action is good in virtue of proceeding from
a certain sort of character, knowledge and pleasure are good
not in virtue of this but in virtue of being, respectively, know-
ledge and pleasure.
If this account of what we mean by "morally good' is correct,
the general answer to our next question, what kinds of thing
are morally good, is clear. Only what is a certain sort of charac-
ter or is related to a certain sort of character in one of certain
ways, can be good in virtue of being a certain sort of character
or of being so related to it. Of the things that are morally good,
I will take for consideration morally good actions, and ask
what can be said about the nature of these. If we could agree
about this, we are not likely to disagree about what sort of men
or what sort of feelings are morally good; they will be the men
that have, and the feelings that spring from, the same sort of
character that morally good actions spring from. Now when
we ask what is the general nature of morally good actions,
it seems quite clear that it is in virtue of the motives that they
proceed from that actions are morally good. Moral goodness
is quite distinct from and independent of Tightness,* which (as
we have seen)' belongs to acts not in virtue of the motives they
proceed from, but in virtue of the nature of what is done. Thus
a morally good action need not be the doing of a right act, and
the doing of a right act need not be a morally good action. The
ethical theories that stress the thing done and those that stress
the motive from which it is done both have some justification,
for both "the right act* and 'the morally good action* are notions
of the first importance in ethics j but the two types of theory
have been at cross-purposes, because they have failed to notice
that they are talking about different things. Thus Kant has
tried to deduce from his conception of the nature of a morally
good action rules as to what types of act are right; and others
have held a view which amounts to saying that so long as our
motive is good it does not matter what we do. And, on the
other side, the tendency of acts to produce good or bad results
has sometimes been treated as if it made them morally good or
bad. The drawing of a rigid distinction between the right and
the morally good frees us from such confusion.
We can perhaps now substitute something more definite for
our preliminary statement that moral goodness is that goodness
which is connected with a certain type or types of character.
If we ask in particular what the moral goodness of morally good
actions arises from, the answer seems (as we have seen) to be
'from a certain kind of motivation*. When we ask what the
moral goodness of morally good feelings arises from, we cannot
answer 'from motivation', for the term 'motive' is inappropriate
here, being appropriate only when action is in question. But
cannot we say that the moral goodness both of actions and
of feelings arises from their-proceeding from a certain kind of
desire ? We think that sorrow at the misfortune of others is
morally good because we think that it springs from an interest
in their happiness, or, to put the matter more plainly, from a
' pp., 4-6.
desire that they shall be happy. As regards actions, the only
objection that might be made to saying their goodness is due
to their proceeding from a certain sort of desire would be the
objection that conscientious action does not spring from desire,
but from something quite different,
Kant, whose greatest service to ethics is the vindication of
the distinct character and the supreme worth of conscientious
action, against the open or covert egoism of most of the eigh-
teenth-century theories, insists that what is operative in it is
something distinct from desire of any kind. It is for him some-
thing so mysterious that he is driven to assign it not to man's
phenomenal nature, which he believes to be moved solely by
desire, but to his real or nouraenal nature, and to hold (as his
theory of freedom implies) that it is not operative in time at all.
Against this it must be maintained that while a general sense
of duty (i, e, an apprehension that there are things we ought to
do) is present, active or latent, in us throughout our adult lives,
it is no more true to say that than to say that a desire for food,
for instance, is present in us, active or latent, throughout our
lives; and that the thought of a particular act as being our duty,
which is what operates when we act from the seme of duty,
arises in us at a particular time just as does the particular desire to
eat, which is what operates when we, in fact, proceed to eat. The
familiar conflict between the sense of duty and other motives
could not take place unless both were operating in the same field,
and that the field of ordinary 'phenomenal' consciousness.
Regarding the sense of duty as an operation of reason dis-
tinct from any form of desire, and accepting the general truth
of Aristotle's dictum that thought alone sets nothing going,1
Kant found one of his greatest puzzles to lie in the fact that pure
reason can become practical, that the mere thought of an act as
one's duty can by itself induce us to do the act. In comparison
with this, actions from desire seemed to him easily intelligible,
But it may be maintained that there is no more mystery in the
fact that the thought of an act as one's duty should arouse an
impulse to do it, than in the fact that the thought of an act as
pleasant, or as leading to pleasure, should arouse an impulse to
fflc. Eth, HJ9» 35.
do it. Human nature being what it is, the latter thought arouses
an impulse to action much more constantly than the former",
but it is only if we have already assumed (on quite insufficient
grounds) that pleasure is the only object of desire, that we shall
find it difficult to suppose that the thought of an act as right
can arouse a desire to do it, Kant himself, of course, allows
that the thought of duty arouses emotion—-the emotion which
he calls Achtung, respect or reverence; and this seems to be
part of the truth. But to hold this is not incompatible with
holding that the thought of duty also arouses desire. Now
people often describe themselves as 'wanting* to do their duty,
It seems to be often a perfectly straightforward description of
their state of mind, and when we examine ourselves we find
something of which this appears to be a correct description.
All that Kant has said of the uniqueness and the moral supremacy
of action arising from the awareness of duty is justified; but it
owes its uniqueness and supremacy not to not proceeding from
desire, but to proceeding from, a desire which is specifically
distinct because it is a desire for a specifically distinct object,
not for the attainment of pleasure nor even for the conferring
of it on others, but just for the doing of our duty. It isin virtue
of the specific quality of the attribute 'obligatory', and not of
arousing something other than desire, that the thought of duty
has the humbling and at the same time uplifting effect that Kant
ascribes to it.
When it is denied that we desire to do our duty, that in us
which leads us to do our duty tends to be thought of as being
in necessary conflict with desire. The sense of duty tends to be
described as the sense that one should do certain acts, though
on other grounds (e. g. on the ground of their painfulness) one
wants not to do them. But "the sense of duty* really means the
sense that we ought to do certain acts, whether or not on other
grounds we desire to do them, and no matter with what in-
tensity we may desire, on other grounds, not to do them. One
of the effects of the forming of a habit of dutiful action is that
any natural repugnance one may have to dutiful acts on other
grounds tends to diminish. If we form a habit of early rising,
jr example, it becomes easier, and less unpleasant, to rise early.
But when the doing of a dutiful action is thought of as neces-
sarily involving a resistance to desire, the paradoxical conse-
quence follows that the forming of a good habit, since it leads
to less and less resistance to desire being necessary, has to be
held to involve that we act less and less from a sense of duty as
the habit grows stronger, so that (if, as Kant holds, sense of
duty is the only good motive, or even if it is the best motive)
to form a good habit is to become a person who acts less well,
and therefore a less good person.
If habitual action could become so automatic that it ceased
to be done from the thought that it is right, there would cer-
tainly be less moral worth in the doing of it. But so long as the
act is still done because it is right, it will have no less moral
worth because it has become easy. The agent's devotion to
duty has not been impaired, but its whole intensity is not
needed, as it was at the beginning, to enable htm to do the right
act. Some of it has now become surplus. The view that there
may be a reserve or surplus of good motive, more than is needed
to produce the right act,1 enables us to avoid the paradoxical
consequence that to form good habits is to become less good.
Goodness is measured not by the intensity of the conflict but
by the strength of the devotion to duty. The stronger the
motives that oppose the doing of the right act, the more sure
we can be that when the right act has been done the sense of
duty must have been strong; but it may well be equally strong
when resisting motives are feeble or non-existent.
Tending, as he does, to think of obligation as thwarting
desire, and holding that a perfect character must have no desires
adverse to the doing of what is right, Kant finds himself obliged
to say that a perfect character or 'holy will* would act not under
a sense of obligation but only under a sense of the goodness or
tightness of certain acts. But, as we have seen, an ac% as dis-
tinct from the doing of it from a certain motive, is nevetgooJ^
and to think of an act as right is to think of it either as being
obligatory or as being one of certain alternatives, the doing of
one or other of which is obligatory;2 so that the distinction
which Kant draws between the right and the obligatory does
' a. p. 172. » cr. pp. 3-4.
not answer to the real difference. When we realize that obliga-
tion is not necessarily in conflict with desire, but merely in-
dependent of desire, we need not hesitate to say that a holy will
would act under a sense of obligation; only it would be an
obligation that would not necessarily be irksome, because there
need not be actually present any desire adverse to it. In fact,
the better a man is, the more intensely will he feel himself
bound to act in certain ways and not in others, though it will
also be true that he will have less desire to act in these other
ways. It would, however, be a mistake to say that all internal
conflict is incompatible with a perfect character. A perfect
character would, ex hypothesi, have no bad desires; it would
not, for instance, desire the pleasure of contemplating another's
pain. But its perfection would not exclude the desire for inno-
cent sensuous pleasures, such as those of food or rest, and these
desires might incidentally be in conflict with the desire to do
what was right. The possibility of such conflict as this seems
inseparable from human nature, and at the same time quite
compatible with a perfect character.
It would seem, then, that conscientious action springs from
a certain desire (the desire to do one's duty), and owes its good-
ness to the specific nature of this desire. And if this be granted,
we have removed the only objection that seems likely to be made
to the contention that all morally good actions (and feelings)
owe their goodness to the kind of desire they spring from.
If 'morally good* means what I have taken it to mean, it
seems that besides (i) conscientious action two other kinds of
action are morally good; and these too owe their goodness to
the nature of the desire they spring from. These are (2) action
springing from the desire to bring into being something good,
and (3) action springing from the desire to produce some
pleasure, or prevent some pain, for another being.
Under (2) I would include actions in which we are aiming
at improving our own character or that of another, without
thinking of this as a duty. And believing as I do that a certain
state of our intellectual nature also is good, I would include
actions in which we are aiming at improving our own intel-
lectual condition or that of others. Some writers would main-
tain that such action, and that in which we aim at producing
pleasure for others, is not morally good, but virtuous. But if
I am right in the account \ have given of what we mean by
moral goodness, viz. that it is goodness due to the character,
or (more definitely) to the desire, involved, it seems clear that
such actions aremorallygood. And it would seem contrary to the
natural usage of the word 'virtuous* to deny that conscientious
action is virtuous. In fact 'morally good' and 'virtuous' appear
to mean the same thing, and to be applicable to actions from
any one of these three desires.
It may seem at first sight doubtful whether the second and
third of these desires are really different. Some might be in-
clined to hold, and I have sometimes myself believed, that to
think of something as being a pleasant state of consciousness
for another being necessarily involves thinking of it as good,
or is thinking of it as good in a particular way, just as thinking
of an action as virtuous is thinking of it as good in another
particular way. There is no doubt that the two thoughts are
often closely associated. Many people undoubtedly do think
that pleasure is good, and in aiming at the production of pleasure
for another are aiming at the production of something thought
of as good. Nevertheless it appears to me that the thought of
something as pleasant for another being is both different and
separable from the thought of it as good,1 For (a) there are
many states of consciousness of other people that we think to
be pleasant and also bad (e. g. states of enjoying the contempla-
tion of some one else's pain). It can only be by overlooking
such cases that any one can come to hold that to think of some-
thing as pleasant for another being either is or necessarily
involves thinking of it as good. And (If) quite apart from such
vicious pleasures, it seems clear that (i) the thought of a state
of consciousness as pleasant for another cannot be the thought
of it as good in a particular way; for if it were, the thought of
a state of consciousness as pleasant for oneself would equally
be the thought of it as good in a particular way (since being
pleasant is, on this view, just being good in a particular way);
On this question, as on some others, I wish to resract the view put forward by
me in PrateeJingi tftkt Aristotelian Sacittjr, 19*8-9,151-74.
and then the desire to get such an experience for oneself would
be the desire to bring into being something thought of as good,
and would itself have to be deemed morally good; whereas we
are convinced that the desire of an innocent sensuous pleasure
(such as that of rest after work) is morally indifferent. And (ii)
it seems clear that the thought of a state of mind as pleasant
either for oneself or for another does not necessarily involve the
thought of it as good, since we should then on reflection always
be able to detect the distinct presence of the second thought
whenever the first'is present; which we certainly are not.
It seems, therefore, that we must distinguish the desire to
bring something pleasant for another person into being, from
the desire to bring something good into being (of which an
instance would be the desire to help a friend to improve his
character).! And both are evidently different from the desire
to do what is right. Yet there is this amount of connexion
between the thought of an act as right on the one hand and the
thought of a good, or pleasure for another, to be brought into
being, on the other hand, that when we think of an act as right
we think that either something good or some pleasure for
another will be brought into being. When we consider our-
selves bound, for instance, to fulfil a promise, we think of the
fulfilment of the promise as the bringing into existence of some
source of pleasure or satisfaction for the person to whom we
have made the promise. And when we consider the other main
types of duty—the duties of reparation, of gratitude, of justice,
of beneficence, of self-improvement—we find that in the
thought of any of these there is involved the thought that what
the dutiful act is the origination of is either an objective good
or a pleasure (or source of pleasure) for some one else. The
conscientious attitude is one which involves the thought either
of good or of pleasure for some one else, but it is a more re-
flective attitude than that in which we aim directly at the pro-
duction of some good or some pleasure for another, since in it
the mere thought of some particular good, or of a particular
This may be held whether or not we consider fhai all pleasures, or that some
pleasures, are in fact ^ood. The desire to bring them into being qua pleasant for another
wili in either case be different front she desire to bring rhem into being qua good.
pleasure for another, does not immediately incite us to action,
but we stop to think whether in all the circumstances the bring-
ing of that good or pleasure into existence is what is really
incumbent upon usj the difference is well illustrated by that
between discriminate and indiscriminate charity.
If there be these three types of virtuous action, it seems that
in principle there might be three corresponding types of vicious
action, (i) the desire to do what is wrong, (2) the desire to
bring into being some particular, evil, (3) the desire to inflict
some pain on another. It seems clear that all of these, if they
exist, are vicious, and that they are different from one another
as are the three corresponding forms of good action. But it is
not quite obvious that any one of them ever really occurs.
(i) Men seem to be attracted to the doing of wrong acts by the
thought of the pleasantness of the particular acts or of their
consequences, or it may be by the thought of the pleasure of
flouting authority and convention; it seems very doubtful
whether they are ever attracted by the mere thought of the
wrongness of an act. 'Evil, be thou my good*, which is (I sup-
pose) meant to indicate this attitude, is the maxim not of a man
but of a devil. (2) Again, when some one sets himself to bring
something evil into existence, e. g. to corrupt some one's charac-
ter, it may be doubted whether it is the badness of what he is
bringing into being that attracts him, or the pleasant sense of
power or some other pleasure that will attend the action. The
corrupting of the character of another, for the sake of corrupting
it, is again the traditional role of the devil; but I am inclined to
think it not impossible for a man. (3) The existence of dis-
interested malevolence seems at first sight doubtful j it may be
argued that what incites any one to do a malevolent act is not
the thought of the pain of his victim, but the thought of the
pleasure he will himself derive from watching or thinking of
the other's pain. But the true view seems to be that the very
thought that that experience will be pleasant to oneself pre-
supposes the existence of an independent desire for the pain of
the victim. And a vindictive death-bed will is pretty good
evidence of disinterested malevolence, as one of the opposite
kind is of disinterested benevolence.
Between the three types of good desire (and of consequent
good action), and similarly between the three conceivable
types of bad desire and action, it seems possible to establish a
scale of goodness, or of badness. For, firstly, it seems clear
that the desire to do one's duty is the morally best motive.
Many will question this, and think that acts springing from
love, from an interest in the well-being of a particular person
or persons, are better than those dictated by the 'cold', 'hard',
and 'rigid* sense of duty. But let us reflect. Suppose that some
one is drawn towards doing act A by a sense of duty, and
towards doing another, incompatible, act B by love for a
particular person. Ex hypotheai^ he thinks he will not be doing
his duty in doing B, Can we possibly say that he will be acting
better if he does what he thinks not his duty than if he does
what he thinks is his duty? Evidently not. What those who
hold this view mean by 'acting from the sense of duty' is obey-
ing a traditional, conventional code rather than following the
warm impulses of the heart. But what is properly meant by the
sense of duty is the thought that one ought to act in a certain
way, not the thought that one has been brought up to or is
expected to act in a certain way. And it seems clear that when
a genuine sense of duty is in conflict with any other motive we
must recognize its precedence. If you seriously think you ought
to do A) you are bound to think you will be acting morally
worse in doing anything else instead. *
Suppose, now, that love and sense of duty incline us to the
same act. Will our action be morally better if we act from the
first motive or from the second ? It seems clear that since the
sense of duty is recognized as the better motive when the two
are in conflict, it is still the better when they are in agreement.
We may like better the man who acts more instinctively, from
love, but we are bound to think the man who acts from sense
of duty the better man. And this is not merely because instinc-
tive affection is a more wayward, capricious motive than sense
of duty, more apt to lead to wrong acts. It is because the
sense of duty is different in kind from, and superior in kind to,
any other motive.
We have considered (a) the case in which sense of duty and
another motive3 present in the same person, point to different
acts, ($) that in which they point to the same act. But there is a
third comparison to be considered. Many people would say
that a man who helps his neighbour from sympathy, without
thinking whether this or any other action is his duty, does a better
action than another man who helps his neighbour from a sense
of duty; though thay might admit that when the sense of duty
arises in a man, it is better that he should act from it than from
any other motive; and this is at first sight an attractive view.
Nevertheless, it appears that action from the sense of duty,
which we have seen to be better than action from any motive
which conflicts or conspires with it, is also better than action
not preceded by the thought of duty. The attractiveness of the
opposite view seems to arise from the following fact: The sense
of duty is more obviously present when it has to fight against
opposing inclinations; i. e. is more obviously present in an
imperfect character than in one all of whose inclinations urge
it in the direction which duty also indicates,1 We are therefore
apt to associate it with an imperfect character, and to prefer the
characters of those whose inclinations are good and who act on
them without thought of duty. But in truth the thought that
an act is my duty is not the thought that an act is my duty
though I want to do something else, but the thought that an
act is my duty whether or not I want to do something else;
though if I do not want to do things incompatible with duty
the act done from sense of duty is done immediately, without
the necessity of a moral conflict and without the thought of
duty having to be held for any length of time before the mind.
The man who acts from a sense of duty without fuss or conflict
appears to be a better character than the man who acts from
inclination without any thought of duty; though, as I suggest
later, the man who acts from both sense of duty and virtuous
inclination is in a still better state.
Next, it seems plain that the desire to bring into being some-
Cf» Kaftt, Grundtegung (Akad. Ausp;. iv, 397—9, Abbott's Kent's Tksory tif Ethicx^
i j—i f}, One fact which makes the presence of the sense of duty more obvious i n the
i ormer cast" is that in this case tltere is present not only the thought that I outfit to do
so-and-so, but also the consequential thought that I ought to resist the inclination
which would prevent me frora doing so-and-so.
thing good is a better desire than the desire to bring into being
a pleasure for another. This is plainly seen in the fact that we
should think concern for the character of one's children,
or friends a better moral state than concern for their
Similarly it seems clear that the desire to bring something
bad into being is worse than the desire to produce pain for
another; just as concern for the character of another is better
than concern for his pleasure, the desire to corrupt some one's
character would seem to be more villainous than the desire to
inflict pain on him. On the other hand, it is less bad than is or
would be (for its existence seems improbable) the completely
anti-moral attitude of desiring to do what was wrong, simply
because it was wrong.
There seem then to be three types of good desire (and conse-
quent action) of diminishing goodness, and three conceivable
types of bad desire (and action) of increasing badness. But it
is clear that most bad actions do not belong to any of these
types, but are selfish actions, in which the desire involved is
none of the bad desires but desire for one's own pleasure. One
is at first sight tempted to say that this desire is in itself morally
indifferent. But it seems that a distinction must be drawn
between good pleasures, bad pleasures, and indifferent pleasures.
By a good pleasure I mean one which springs from the satis-
faction of a good desire, as e. g. the pleasures of a good con-
science spring from the satisfaction of the desire to do right,
or the pleasure of doing a kind action springs from the satis-
faction of the desire to produce pleasure for another. By a bad
pleasure I mean one which springs from, the satisfaction of a
bad desire, as e, g. the pleasure of making another suffer springs
from the satisfaction of the desire to make him suffer. By an
indifferent pleasure I mean one which does not spring from
either of these sources; and it is plain that most selfish actions
are dictated by desire for such pleasures, as a rule for the pleasures
of the senses. It seems clear that the desire of an indifferent
pleasure is in itself indifferent; that the desire of a good pleasure
is either indifferent (as I am inclined to think) or less good than
any of the three good desires previously named; and that the
desire of a bad pleasure Is bad, but less bad than any of the three
bad desires previously named.
The general character of selfish action, in distinction from
the three forms of bad action first mentioned, is that it proceeds
not from a bad, but from an indifferent desire. When one asks
on what its badness depends, one is at first sight tempted to say
that there is some badness in feeling any desire more strongly
than one feels another desire higher in the scale. But this clearly
will not do. Take, for instance, a soldier whose duty it is to
rest before the battle of the following day. He may be so tired
that the desire for the morally indifferent pleasure of resting is
much stronger in him than the desire to do his duty; and no one
would think him selfish or think any the worse of him for that.
It is, therefore, not the mere feeling of a lower desire with
greater intensity than a higher that constitutes selfishness and is
morally bad. It is only when the action aimed at securing
pleasure for oneself excludes and makes impossible the doing of
one's duty or the bringing into being of something that is good,
or some pleasure for another, that the action in question is
deemed to be morally bad. Such action seems to be of two
types, or rather (since these probably shade off into one another)
to range between two extremes. At one extreme, it is the
choosing of one's own pleasure in deliberate preference to the
doing of duty or to the bringing into being of a good, or of a
pleasure for another. At the other, it is the doing of an act
which in fact excludes the doing of duty or the bringing of some
good or some pleasure for another into being, though the
thought of a duty that should be done, or of a good, or of a
pleasure for another, that might be brought into being, is not
present to the agent's mind. The two types might be called
deliberate and impulsive selfishness respectively; of the two,
deliberate selfishness is the worse. But even what might be
called impulsive selfishness usually includes the vague aware-
ness that if one stopped to think one would see that, in doing
what one does, one is omitting an opportunity of doing one's
duty, or of bringing some good, or some pleasure for another,
into being.
There are these two main kinds of bad actions—a simpler and
worse kind which consists in acting on a bad desire, and a more
complicated and less bad kind which consists in acting on an
indifferent desire to the exclusion of an action to which a good
desire prompts us, or would prompt us if we stopped to think.
One thing which I have taken for granted may at first sight
appear somewhat paradoxical, viz, that while the desire to
bring something good into being is good whether the good
thing is to be a state of oneself or of another (e. g. an improve-
ment in one's own character or in that of a friend), the desire to
bring a pleasure into being for another is good, but the desire
to get a pleasure for oneself is indifferent when the pleasure is
indifferent (perhaps we should add *or when it is good*), and
bad when the pleasure is bad. A complication must be
introduced here which I have omitted from my main
statement in order that its general outline might be clearer:
viz, this. The distinction between good, bad, and indifferent
pleasures may be applied to the pleasures we desire to pro-
duce for others as well as to those we desire to get for our-
selves,* and when we apply it we see that the desire to secure a
bad pleasure for another (e.g. the pleasure of watching some
one suffer) is itself bad. Still, the main contrast is clear, that
the desire to produce an indifferent pleasure for another is good,
and the desire to produce it for oneself indifferent. That this is
what we really think seems perfectly clear. And it is not really
paradoxical. For though the pleasure will be similar whether
it is enjoyed by the agent or by another, the desire for pleasure
for oneself is quite a different thing from the desire to produce
it for another, since one's relation to oneself is entirely different
from one's relation to any other person.

So far I have spoken as if each action sprang from one and

only one motive; we have next to consider actions (if there are
any) done from mixed motives. Is it possible to do an act
partly from sense of duty, partly from love, and, if so, is such
an action morally better or worse than one done from pure
sense of duty? Again, is it possible to do an act partly from a
good, partly from an indifferent or bad motive, and, if so, what
are we to say of the moral value of such an action?
It is clear that a mother, for instance, may want her child to
be happy or good, and may also think it her duty to produce
for it the conditions of happiness or of goodness. But can she
really act from both motives together; must not one be opera-
tive and the other ineffective? The metaphor of 'mixture of
motives* seerns a questionable one, but we must at least recog-
nize the co-operation of motives, or perhaps we should rather
say 'the co-operation of elements to form a single motive'. We
are familiar, I think, with cases in which conscience alone or
self-interest alone would not have induced someone to do some
difficult act, but the two together have induced him to do it. The
activities of most statesmen are probably correctly attributed
to the co-operation of ambition, party feeling, and patriotism,
in varying proportions; and it does not seem as if we can always
say, 'This act was due to ambition and that to party feeling and
that to patriotism*; in many of their acts it seems that two, or
perhaps all three, of these motives are at work. Again, if we
want to induce a man to undertake a certain task, we probably
put various arguments before him representing the points in
its favour, and we do so not merely in the faith that if one point
does not appeal to him another will, but in the belief that the
various points, or some of them, may have a cumulative
influence on Mm; and this seems, in fact, often to be the
Can we go further, and say that all motives that are present
at all in a man's mind at a particular moment are bound to co-
operate in producing his action, the mind being subject to a
law of composition offerees of its own? This is clearly not
the case. If motive A inclines us towards a certain act and
motive B towards a different act, our doing of the first act
is not necessarily altered (though the state of feeling with
which we do it is) by our desire to do the second, nor do we
necessarily do an act intermediate in character between
the two. The doing of such an act (e. g. when there are two
safe ways across a street, and we take an unsafe way between
them) is a mark of distraction and loss of self-control. When
we act with self-control we do the one act resolutely, irres-
pective of a lingering wish to do the other. We must, how-
ever, consider separately the case in which two motives point
to the same act. We may act from both motives together;
must we do so ? We have seen that a motive adverse to the
doing of a particular act may be so disregarded that the act is
done just as it would have been done had the motive been
absent; it seems reasonable to suppose that a motive favour-
able to the doing of the act may equally be disregarded; e. g.
that when both sense of duty and ambition would incline a
man to do an act, he may, though he is sensible of both motives,
do the act wholly from the one or wholly from the other;
and attention to one's own experience seems to support this
Where, however, a higher and a lower motive do co-operate
in inducing us to act, what degree of moral worth has the action
—more, or less, than if it had been done from the higher motive
only? Kant assumes that its worth will be less, since it will no
longer be 'pure'. The question can be looked at in either of
two entirely different ways. It will suffice if we lake account
of two morally good motives, one higher than the other, of
one morally indifferent motive, and of one bad motive; and
must suppose that each of these motives has a definite
amount of effectiveness in inducing the agent to act. We are
far from being in a position to assign correctly definite numeri-
cal amounts either to the value or to the strength of motives;
but it will help us to attack the problem if we assign (admittedly
arbitrarily) such amounts. Let us suppose, then, that in the
scale of values sense of duty is represented by 10, love by 8,
desire of an innocent sensuous pleasure by o, malice by — 8:
and let us suppose an act in which two of these motives co-
operate with equal strength in producing an act. To assign
other values to the motives, or unequal strengths to them,
would not affect the argument. Kant, in insisting that the worth
of an action is degraded by the presence in it of any motive
lower than the sense of duty, is really assuming that its value
must then be the value half-way between that which it would
have had if done wholly from sense of duty, and that which it
would have had if done wholly from the lower motive; e. g. an
act done from sense of duty+desire of an innocent sensuous

r will have the value An action done from

sense of duty+malice1 will have the value . And since

Kant assigns no value, positive or negative, to 'pathological*
love, an action done from sense of duty+love will have the
value . If we amend his view, as we have seen reason to
do,2 by assigning a positive value (say 8) to love, an action done
from sense of duty+love will have the value
It is, however, possible to take quite a different view; to hold
that, for instance, sense of duty may be present with undimi-
nlshed strength when another motive co-operates with it, and
that when it is, the addition of a morally indifferent motive does
not lessen the value of the action, and the addition of a good,
though less good, motive increases it. The action done from
sense of duty+desire of an innocent sensuous pleasure may be
estimated at 10+0; that done from sense of duty+ malice at
10—8; that done from sense of duty + love at 10+8. If we call
sense of duty A, love B, desire of innocent sensuous pleasure
Ct malice D, the value of the actions we have been considering
will be estimated thus:
Kant. Kant modified, alternative view.
Action from A 10 10 10
„ „ 4+B 5 9 18
» » 4+C 5 5 10
„ „ A+D i i z
It is now, perhaps, easier to see the principles on which the
two modes of estimation are based. In insisting (as in principle
he does) that when two motives are equally effective, their
total (or net) value must be divided by two in order to ASSEESS
the value of the action, Kant is really assuming that all similar
actions have an equal total intensity of motivation, so that by
* This combination may sound paradoxical; but a little reflection will show that it
is not impossible, nor, indeed, at all uncommon,
* Cf. pp. 160-1.
the extent to which any other motive is effective, the sense of
duty must be less effective than it would have been if acting
alone. We, on the other hand, are suggesting that the sense of
duty may be present with equal intensity and effectiveness
whether another motive is or is not effective. Or, to put it
otherwise, he is assuming that the motivation is always exactly
enough to produce the doing of the given act, while we are
suggesting that there may be some to spare, so that in a man
who does an act, e. g., from sense of duty+love, the sense of
duty may yet be strong enough to have secured the doing of
the act from it alone; i. e., while we agree with Kant that there
are cases in which, say, sense of duty+a morally indifferent
motive are just strong enough to produce a certain act, and that
then the value of the act, if both motives are equally effective,
is only half what it would have been if done from pure sense of
duty, we suggest that there are other actions in which, while a
morally indifferent motive is present, the sense of duty is strong
enough to have secured by itself the doing of the act, and that
in that case the action is as good as if it had been done from
sense of duty alone. And while we hold that there are actions
in which sense of duty+an inferior good motivearejust enough
to secure the doing of the act, and that in that case the value of
the action is reduced, we suggest that there are other cases in
which, though an inferior good motive is present, the sense of
duty is strong enough to have secured by itself the doing of the
act, and that in such a case the value of the action is greater
than if it had been done from duty alone. On the question
whether such a reserve or surplus of motivation exists, there
seems to be no appeal except to introspection, and it seems
clear on introspection that there are many cases in which, while
a variety of motives inclines us to do an act, one of them or
some of them would have been enough to induce us to do it.
Thus we can, while agreeing with Kant that the sense of duty
is the best motive, justify the generally entertained preference
for actions in which some more instinctive generous impulse
is present as well. And experience suggests that the presence
and effectiveness of instinctive generous emotions are by no
means adverse to the operation of the sense of duty. It is not
the case that men in whom the sense of duty is strong are usually
less affected by the generous emotions than those in whom it
is weak. And it is possible, as we have tried to show, to value
highly the presence and operation of warm personal feeling,
without disparaging, as it has so often been thought necessary
to do, the supreme moral value of the sense of duty.
This page intentionally left blank
5 that *l ought' implies 'Ican''. Although Ross does not state It here, he
thinks that tightness is not constrained by ability in this way. For
example, he maintains thai 'we can still call grief the right or fitting
emotion in certain situations.., even if we do not suppose the person
we are dunking about has it in his power to feel grief in those circum-
stances' ( The Foundations of Ethics (Clarendon Press: Oxford, 1939),
55). Ross does, however, go on to say that 'it must be granted that
when we use 'right' o:i acts, as opposed to emotions, we usually think
of them, as being in the agent's power* ( The Foundations of Ethics, 55).
But the implication here seems to .be merely conversational rather
than the strict, entail merit he thinks holds between *I ought' and 'I
7 the tiling done: In The Foundations oj Ethics Ross changed his mind
about this. He there thinks that what we should do is not certain acts,
but 'set ourselves* to do certain acts. Following H. A. Prichard
('Duty and Ignorance of Fact*, in Prichard, Moral Obligation, ed,
W. D. Ross (Clarendon Press: Oxford, 1949) ), he wrote: "contrary
to the implication of ordinary language and of moral rules in partic-
ular., an obligation must be an obligation, not to do something, but to
perform an activity of a totally different kind, that of setting or exert-
ing ourselves to do something, i.e. to bring something about *, On his
later view, therefore, success in doing what one sets oneself to do is
irrelevant to whether one has succeeded in doing what one should.
This contradicts what he says later in The Right and ike Good about
success (42—7),
7 from which it is done: In The Foundations of Ethics Ross modified his
view slightly- He maintained that the moral goodness of an action is
determined not only by the agent's actual motives, but also by the
motives he would have if he were responsive to all of the morally rel-
evant considerations in the right proportion (Poimdations^oCt—iQ),
See note top. 156,
7 our moral thought". In The Foundations of Ethics Ross makes it clear
that he does not suppose that acts are independent of motives in the
sense that the former could exist without the latter. All he is keen to
maintain here is that the tightness of acts is independent of the
motives from which they are done, and so should be considered in
abstraction from their motives (Foundations, 127—8).
20 whole nature: Although Ross does not tell us here why he chose file
misleading term 'prima. jade dirty*, he dyes in The Foundations of
Ethics. First, this term, picks out the fact: that when we consider
what: we should do in particular situations it. is the fitness or unfit-
ness oi an imagined act in certain respects that first catches our
attention.',,, it is only by recognizing these different features one
by one that we can approach the forming of a judgement on the
totality of its nature; our first look reveals these features in isola-
tion, one by one; they are what appears prima facie' (Foundations,
84). Secondly, it picks out the deontic nature oi these features
the fact that they make certain actions in some respect right or
wrong, or better, fitting or unfitting to different aspects of the situ-
ation (ibid.),
20 a claim on oneself. Although Ross rejected Prichard's suggestion of
the term 'claim', he later accepts Carritt's suggestion of the term
"responsibility* (The Foundations of Ethics, 85), He also seems
happy to use the term 'fitting" or "suitable * as a good substitute for
'prima.facie duty* (ibid. 51—5 and 79— 82, especially pp. 53 and 84),
His view is, then, that to be prima facie right is to be fitting to some
aspect of the situation—that is, to sonic responsibility we have. To
say that an act is our duty proper is to say that it is the one which is
most fitting to all of the relevant features of the situation, that it is
the one that fulfils as many of our responsibilities as is possible, A
right act which is nonetheless prima facie wrong is one that fulfils as
many of our responsibilities as is possible, and is thus the act that is
most: fitting to the situation, though it fails to fulfil some responsi-
bility, and is thus unfit to an aspect of die situation,
26 in this way: In The Foundations of Ethics Ross abandoned the view
that we have a prima fade duty to promote our own. innocent plea-
sures (pp. 272—4), In fact he goes so far as to say: 'That we are con-
scious of no duty to maximize pleasure for ourselves seems to be so
clear as not to need argument' (p, 273). He makes no reference here
to his earlier view in The Right and the Good,
30 need no proof'. In a paper published three years before The Right and
the Good Ross maintains that self-evident principles require no
proof, but rejects the view that they cannot be proved. He writes,
"the fact that something can be inferred does not prove that it cannot
be seen intuitively' ('The Basis of Objective Judgements in Ethics*,
InternationalJournalof Ethics, 37 (1927), 121). If he thinks that soine
proposition can be inferred from (justified by) other propositions
and be self-evident, he clearly thinks that its being self-evident does
not rule out the possibility of a proof,
32 morally good action: In The Foundations nj Ethics Ross completely
changed bis mincf on flu's topic. There he argues that what we should
do is not determined by the objective facts—certain relations in
which we stand to others but by certain subjective facts, namely,
what we believe is right or morally suitable in the circumstances as
we take them to be (Foundations, irti—4).
.131 not vice versa: In The Foundations oj Ethics (pp. 254—5) Ross adds
that evaluative judgements express our favourable attitude towards
the object, but notes that this does not imply that what we are
asserting of the object is that it is the object oi such a favourable,
135 is good in itself: In The Foundations of Ethics Boss abandoned the
view that pleasure is intrinsically good (Foundations, 283). He still
maintained that: it is good, but argued that it is good in a relational
sense that is, in the sense that it is a worthy object of sympathetic
satisfaction (Foundations, 275 6).
150 with infinite duration: IE The Foundations of Ethics Ross realized that
the view he expressed here must be wrong. He notes that 'if virtue
were really on the same scale of goodness as pleasure, then pleasure
of a certain intensity, if enjoyed by a sufficiently large number of
persons or for a sufficient time, would counterbalance virtue pos-
sessed or manifested only by a small number or only ior a short rime'
(foundations, 275), Since he could not accept that pleasure is ever
better than virtue, he rejected the view that they are on the same
evaluative scale, and maintained that pleasure is good in a different
sense from that in which virtue is good (ibid,). See note to p, 135.
156 independent of lightness: In The Foundations of Ethics Ross weak-
ened his claim about the independence of moral tightness and
, , , an action "will be completely good only if it manifests the whole
range of motivations by which an ideally good man would be affected
in the circumstances, a sensitiveness to every result for good or for evil
that the act is foreseen as like!v to have, as well as to any special prima
facie obligations or disobligations that may be involved; and only ii it
manifests sensitiveness to all these considerations in their right propor-
tions. But if the agent is responsive to all the morally relevant consider-
ations in their right proportions, he will in fact do the right act. Thus no
action will fsave tiie utmost moral excellence whicli an. action in the cir-
cumstances can have, unless it is also die right, action, (foundations,
164 any thing else instead', Ross never abandoned the view that the sense
of duty is the best motive {The Foundations of Et/tics, 303—5). But
because he came to believe that the moral worth of an action
depends not only on one's actual motive, but also on motives one
would have if one were appropriately responsive to all the moral
considerations in the situation (see note to p , 156), he came to
believe that an action done from a sense of duty could be morally
The nepotist may act from the thought that he ought to bestow the job
on B; and so far there is an element of goodness in his action. But in so
far as he is failing to be influenced by the thought that he ought to do
j ustice to C, D, and the rest, his action is a bad one; and it is easv to imag-
ine a case in whicli theprima, facie obligations he is failing to be influ-
enced by are much more weighty than those he is being influenced by;
and in. such a case his action will be on die whole positively bad,
(Foundations, 307—8)

"Act' and 'action' 7 and vice 163

Animals: Durkhekn, H, 12—15
duties to 48-50 Duty:
pleasures ol 157—8 to animals 48—50
rights of 50,52,54 and blame 32
Aristotle 42,69, 157 conflict of 18 19,13,28
Asceticism 552 correlaiivity of duties and rights
48 0
Beauty 69- 71, 7$ 7, 89,116,118-', i distinguished from, right 3—4
Bentham, J, 7 -8,144 5 independent of, but not opposed
Bradley, F, H. 57-147 to desire 159—60
judgements about 29—33
Claims ^o not based on available evidence 52
Colour 86-8,104 5,115, 119 not to aim at some end 4 3
Commensurability of goods 142 54 not to produce a change in body
Conflict of duties 18—9,13, 28 or mind 42
Croce, B. 107—9 o t perfect obligation 18
personal character of 22
Definition: and relations to oihers ic>
its species 1,6—7 to secure outcome 43—45
method for assessing definitions sense of 5 6, 157—60, 164—5,
of'good' 93- -4 168- 73
Desire: success, criterion of performance
to bring pleasure to self and to of 45
others, value of 168 utterly independent of
conflict of, inelinrinable \ 60 consequences /sj—47
to do one's duty the best motive
6, 164-5 Egoism 16
andduty 159—60
to give pleasure and promote Feeling and value judgements 131—2
good distinct 161—2
and goodness 95—100,156 61 'Good':
and good pleasures 166 adjunctive/aitributive use of
die nature of our desire, the basis 65-7
of moral goodness and badness commensurability of goods
116—64 142-54
the. non-existence of the object of disagreement about the 83—4,95
96 justice, intrinsically good 138
to produce good better than desire knowledge and belief intrinsically
to produce pleasure 165—6 good 138—40
Good : (COT i and secondary qualities So
the meaniung of 65-74 a tori-resultant property 104- -5,
method tor j ^_ % ,m , Jtlii 'lions 12i 3
of 93 4_ ultimate 68—9,73—4
as object of interest 8o-j, 85—104 and 'what ought to exist 105 6,
pleasure intrinsically good 15 5 • • -8 no -11
predicative use of 64,67 Greene, T. H. 50—52
value judgement and feeling
13 i—z Hedonism 10,99, 144,145
virtue intrinsically gcsod 134 5. Hobbes, T. fi, 56
what tilings are good? 134-41
Goodness: Indehnability:
analogy with colour 86—8, 104—5, how established 91—4
115,119 of goodness 81, 86, 87-8, 91-4
the coherence theory 78—9 Intrinsic goodness 68—9,74,75,
a consequential attribute 79, 88, 1:15—22
contributive 72- 3 justice .it, 26-7.5>-j, 138
degrees of Mi-54
and desire 95—100 Kant, I. 5,18,16,87,109,136,
the fuliiilment-ol-tendency 156—9,165,170-2
theory 79-80 Knowledge:
indefinability/simplicity of 82, eommensuration of different states
86,87-8,91-4 of, and right opinion 145-9
instrumental 68 comtnensuration of knowledge
intrinsic 68—9,74,75, 115—112 with other goods 149- >o,
Prof. Moore's study of intrinsic 151-3
goodness 114-an contrast widi belief 146—7
moral 66,101—2,155-7} Inferential and non-inferential
the nature of 75-r 3-5 knowledge 148
the objective- (Objectiv-) theory intrinsic goodness of 1^8-40:
104-14 not a species of the same genus as
prima.facie goodness of pleasure belief 146
137-8 value of knowledge depends in
property only of mental states 86, part on. the nature of fact appre-
122,140-1 hended 146—7
purely objective 131-1 and welfare 148
view that it is not a quality reject-
ed :io<>—114 Laird, Prof. |. 88
psychological or subjective theo- Law arid promises 63—4
ries 80 104 Levy-Bruhl, Prof.!_,, 12 15
relational theories 75—104
it relational then no intrinsic McTaggart, J. M. E. 113,144
goodness 77—8 Meinong, A. 10411
Merit 135-8 Plato tf>
Mill, J, 8. 8,145,149 Pleasure;
Moore, Prof. G. E. vi, t, 6~i i, of animals 137-8
16—19,14; 57* 68-73,75,82, duty to promote our own pleasure
85-8,91-4 24-6
Moral code; good, bad and indifferent
and objective truth 14 pleasures 136—8, ( f t , 166—38
origin and validity 14 interior in value to virtue and
Moral convictions of plain man knowledge 149—50
20-1.39-41 intrinsic goodness of 135-8
"Morally good': not commensurable with virtue
distinguished from 'right' 3—6, s 50— i
11.2, 1.56 pnrna, facie g'Oodness of 137—8
moral goodness 15 5 73 relative value of pleasure i.4i-~45,
moral goodness defined 134,155 '49 54
moral goodness and desire and right 162—3
156— IfJl thought of something as pleasant
moral goodness infinitely better not the same as thought of it as
than knowledge 152 3 good ]<5i-3
morally good character consistent Power and rights 52,56
with internal conflict 160 Pilchard, Prof. H. A, v, 20
'morally good' synonymous with Prima fade duty;
Virtuous' i<Si actual duty has greater balance of
only actions done from good pnma facie tightness Qvttpnma
motive are morally good 4,156 facie wrongness 41,46
Motive: analogy arid disanalogies with
duty to cultivate good 5,6n natural laws 2-8—9
mixed motives 168-73 basis ofprima fade duty of non-
not included in what is my duty maleficence 26
4-6, Z2-3 basis viprima facie duty of obey-
sense of duty the best 6,164—5 ing the law 17—8
value of mixed motives 170-73 basis ofpritna facie duty of repara-
tion and gratitude 27
Objective, is goodness an? 104—14 of beneficence and self-
Organic unities 57,69—73 improvement rest on the same
Open question argument 85, 92,94 ground 26
Opinion, degree of value of 146—7 as conditional duty 19
Origin and validity 12—15 distinguished from 'duty proper*
'ought' implies 'can' principle 5 89-20,28
division of 21—1
Pain when deserved, good 130 foundations ofprima facie duties
Parodu, D. 15 19—20
Perry, Prof. R. B, 75-8, 80-2, of injuring wrong-doers 63,
85-104 64
Prttna facie duty: (cant.) and promises 63-4
01 justice grounded in promotion promoting the general interest. 63
of the good 2.7 proportionality 62
of justice reinforced by prima facie retribution 57—60
duty of reparation 28 severity determined by ilieprima
and moral conflict $8--19, 23, 28 fade duties of injuring wrong-
no principle tor resolving conflict doers and utilitarian theory of
23, 56,60
not a duty zo
not a merely apparent duty 20 Qualities, constitutive or fundamen-
as parti-resultant attribute 38, 33 tal, and consequential and
principles oipnma facie duty resultant 104—5, 12I ~ 2
known, actual duty not 29—31
problems with the prima fade duty Rashdail.Dr. H. 18
to produce pleasure for oneself 'Right':
24-6 development of our notions of it
of promise-keeping and
consequentialism 34 "3*5 difference between it and 'morally
to promote the good 58 good' v-6,156
to promote the general interest difference between it and 'my
6y, 64 duty* 3-4
of reward and punishment 58 disanalogy with mathematical
and self-evidence 31- 3 properties 33 4
to spare the innocent 61 does not mean 'ordained by any
as tendency to be our duty 18n, given society' 12
28 does utilitarianism claim to define
term misleading 20 it? 7-«
terra 'claim' rejected 20 evolutionary theory un
unsystematieity, charge of 23 ideal utilitarianism criticized
prima facie wrongoess of lying 16—39
2 [,54-5 its meaning 1—15
P rogress, moral 12 Prof, Moore's claim to define it
Promises, keeping of 17—19, 21,22, 8-11
?5 34-5- 37-40,42-7,63-4 nature of right acts 32,41—7
Punishment: refers entirely to the thing done 7
affixing and inflicting die penalty subjective definitions oi it u
61 what makes right acts right?
criminal desert 57 16-47,123,132-3
deterrence and reformation 56 Rights:
as means of protecting rights animal rights 50,52,54
60-1 correlative with duties 48—56
Moore's view 57 legal 5;o--i, 52
problems with retribution theory natural 5 5—6
58-60 and power 52,56
and punishment. 59—63 Simplicity/ inif efinahility of
purpose of the state, to protect goodness 82,86, 87 8,91-4
rights 59-60 Slavery 51
right to beneficence 5 2l—3 Sociological school 12— i $
right to punish 57 inconsistency f;f 1.3
Ritchie, D, G. 49 sociological enquiry irrelevant tc
Russell, B. 104 ethics 15
Socrates 27
Secondary qualities 86 Special obligations 17,47
Selfish action 166 7 Spinoza, B, 51
actual duty not self-evident 30 Urban, Prof. W, M. So, 104-14
analogy with knowledge of math- Utilitarianism:
ematical propositions 3.1-3 does it claim to define "right'?
apprehension of self-evident T-- H

propositions, requires maturip ideal utilitarianism criticized

12,29,32 16-39
and certainty jo ideal utilitarianism, ignores the
coextenssveness of the right and personal character of duty 12
the optimific not sei ('-evident ideal utilitarianism oversimplifie;
?<i our relations to others i<j
how moral principles become its relation to hedonistic utilitari-
self-evident to us 32-3 anism 16-17
nature of* 12, 29 no self-evident connection
oipruna faa'e duties 12, 20—1 n, between 'right* and "optirniftc'
V 3 .\">"6.
stringency oiprima-jade dunes, utilitarianism in general criticize!
not self-evident 41 43 7
arid trust in reason 30
and 'what we really think' 59—41 Vice, and desire 163
Sense of duty: Virtue:
the best motive 6, (64—5 the best thing 153
and desire 157—60 intrinsically good iV4—5>
different in kind from other value relative to other goods
motives 164 149-54
and mixed motives 168—73
no duty to act from 5—6 Welfare and knowledge 148
Sheldon, Prof. W, H. 80 Wilson, J. Cook 82,92,118,
Sidgwick,H. 7-8 839-40